STREET-LEVEL

BUREAUCRACY
DILEMMAS OF
THE INDIVIDUAL
IN PUBLIC SERVICES
MICHAEL LIPSKY
Street-Level Bureaucracy
STRE ET-LE VEL
BUREAUCRACY

Dilemmas of the Individual

in Public Services

MICHA EL LIPSK Y

Russell Sage Foundation

NEW YORK
 The Russell Sage Foundation

The Russell Sage Foundation, one of the oldest of America's general pmpose founda-
tions, was established in 1907 by Mrs. Margaret Olivia Sage for "the improvement of
social and living conditions in the United States." The Foundation seeks to fulfill this
mandate by fostering the development and dissemination of knowledge about the coun-
try's political, social, and economic problems. While the Foundation endeavors to assure
the accuracy and objectivity of each book it publishes, the conclusions and interpreta-
tions in Russell Sage Foundation publications are those of the authors and not of the
Foundation, its Trustees, or its staff. Publication by Russell Sage, therefore, does not
imply Foundation endorsement.

Kenneth D. Brody
W: Bowman Cutter, III
Robert E. Denham, Esq.
John A. Ferejohn
Larry V. Hedges
BOARD OF TRUSTEES
Mary C. Waters, Chair
Kathleen Hall Jamieson Shelley E. Taylor
Lawrence F. Katz Richard H. Thaler
Melvin J. Konner Eric Wanner
Sara S. McLanahan
Nancy Rosenblum

Library of Congress Cataloging-in-Publication Data

Lipsky, Michael.
Streel-level democracy : dilemmas of the individual in public services I
Michael Lipsky.-Updated ed.
p.cm.
Includes bibliographical references and index.
ISBN 978-0-87154-544-2 (alk. paper)
l. Social workers. 2. Social policy. I. Title.
HV4l.L53 2010
361.301--dc22 2010003955

Copyright 2010 by Russell Sage Foundation. All rights reserved. Printed in the United
States of America. Original edition published in 1980. No part of this publication may
be reproduced, stored in a retrieval system, or transmitted in any form or by any means,
electronic, mechanical, photocopying, recording, or otherwise, without the prior written
permission of the publisher.

Reproduction by the United States Government in whole or in part is permitted for any
purpose.
The paper used in this publication meets the minimum requirements of American Na-
tional Standard for Information Sciences-Permanence of Paper for Printed Library
Materials. ANSI Z39.48-1992.
RUSSELL SAGE FOUNDATION
112 East 64th Street, New York, New York 10065
10987654321
For Joshua and Jacob
 CONTEN TS

PREFACE: DILEMMAS OF THE INDIVIDUAL IN

PUBLIC SERVICES xi
ACKNOWLEDGMEN TS xxi

PART I
INTRODUCTI ON

1. The Critical Role of Street-Level Bureaucrats 3

CONFLICT OVER THE SCOPE AND SUBSTANCE OF
PUBLIC SERVICES 4
CONFLICT OVER INTERACTIONS WITH CITIZENS 8
2. Street-Level Bureaucrats as Policy Makers 13
DISCRETION 13
RELATIVE AUTONOMY FROM ORGANIZATIONAL
AUTHORITY 16
DIFFERENCES BETWEEN STREET-LEVEL
BUREAUCRATS AND MANAGRS 18
RESOURCES FOR RESISTANCE 23

PART II
CONDITIONS OF WORK
INTRODUCTION 27
3. The Problem of Resources 29
DEMAND AND SUPPLY, OR WHY RESOURCES ARE
USUALLY INADEQUATE IN STREET-LEVEL
BUREAUCRACIES 33
4. Goals and Performance Measures 40
GOALS 40
PERFORMANCE MEASURES 48
5. Relations with Clients 54
NONVOLUNTARY CLIENTS 54
CONFLICT, RECIPROCITY, AND CONTROL 57
THE SOCIAL CONSTRUCTION OF A CLIENT 59
vii
Contents
6. Advocacy and Alienation in Street-Level Work 71
ADVOCACY 72
ALIENATION 75
IMPLICATIONS OF ALIENATION 79

PART III
PATTERNS OF PRACTICE
INTRODUCTION 81
7. Rationing Services: Limitation ofAccess and Demand 87
THE COSTS OF SERVICE 88
QUEUING 95
ROUTINES AND RATIONING 99
8. Rationing Services: Inequality in Administration 105
A COMMENT ON THE UBIQUITY OF BIAS 111
9. Controlling Clients and the Work Situation 117
HUSBANDING RESOURCES 125
MANAGING THE CONSEQUENCES OF ROUTINE
PRACTICE 133
10. The Client-Processing Mentality 140
MODIFICATIONS OF CONCEPTIONS OF WORK 142
MODIFICATIONS OF CONCEPTIONS OF CLIENTS 151

PART IV
THE FUTURE OF STREET-LEVEL
BUREAUCRACY
11. The Assault on Human Services: Bureaucratic
Control, Accountability, and the Fiscal Crisis 159
HOLDING WORKERS TO AGENCY OBJECTIVES 162
ACCOUNTABILITY AND PRODUCTIVITY 170
STREET-LEVEL BUREAUCRATS AND THE FISCAL
CRISIS 172
12. The Broader Context of Bureaucratic Relations 180
CONTRADICTORY TENDENCIES IN STREET-LEVEL
BUREAUCRATIC RELATIONS 188
viii
 Contents
13. Support for Human Services: Notes for Reform and
Reconstruction
DIRECTIONS FOR GREATER CLIENT AUTONOMY
DIRECTIONS FOR CURRENT PRACTICE
THE PROSPECTS AND PROBLEMS OF
PROFESSIONAL ISM 201
KEEPING NEW PROFESSIONAL S NEW 204

14. On Managing Street-Level Bureaucracy 212

AN EVOLVING POLICY ENVIRONMENT FOR STREET-LEVEL

BUREAUCRACY 212

SHAPING STREET-LEVEL BUREAUCRATS ' PERFORMANCE 221

INVESTING IN STREET-LEVEL BUREAUCRATS 229
CONCLUSION 236

NOTES 239

m~ ~

ix
 PREFA CE

Dilemmas of the Indiv idual in

Public Services

Prepared For The Thirtieth Anniversary, Expanded Edition

Of Street-Level Bureaucracy

This book is in part a search for the place of the individual in those public
services I call street-level bureaucracie s. These are the schools, police and
welfare departments , lower courts, legal services offices, and other agencies
whose workers interact with and have wide discretion over the dispensation
of benefits or the allocation of public sanctions.
It is also an inquiry into the structure of one of those resonant moments
in civic life. Like driving on interstate highways, playing in a public park,
voting, dining in a smoke-free restaurant, paying taxes, and listening to Na-
tional Public Radio, interactions with street-level bureaucracie s are places
where citizens experience directly the government they have implicitly con-
structed. Unlike these other experiences, however, citizen encounters with
street-level bureaucracie s are not straightforward; instead, they involve
complex interactions with public workers that may deeply affect the bene-
fits and sanctions they receive.
Street-Level Bureaucracy was originally published in 1g8o and made two
distinctive claims. The first was that the exercise of discretion was a critical
dimension of much of the work of teachers, social workers, police officers,
and other public workers who regularly interact with citizens in the course
of their jobs. Further, the jobs typically could not be performed according
to the highest standards of decision making in the various fields because
street-level workers lacked the time, information, or other resources neces-
sary to respond properly to the individual case. Instead, street-level bureau-

xi
 Preface
crats manage their difficult jobs by developing routines of practice and psy-
chologically simplifying their clientele and environment in ways that
strongly influence the outcomes of their efforts. Mass processing of clients
is the norm, and has important implications for the quality of treatment and
services.
These observations are instructive in themselves, and have profound im-
plications for public policy. They suggest that understanding public policies
in street-level bureaucracies requires analysis of how the unsanctioned work
responses of street-level bureaucrats combine with rules and agency pro-
nouncements to add up to what the public ultimately experiences as agency
performance.
The second claim was that work as diverse and apparently unrelated as
that of guidance counselors, judges, police officers, and social workers to a
degree is structurally similar, so that one could compare these work settings
with each other. Describing front-line public service delivery in terms of a
small number of analytic characteristics made possible a new way of seeing
these very familiar public roles, and how they were like and different from
one another.
However diverse these occupations otherwise are, they could now be
seen as embodying an essential paradox that plays out in a variety of ways.
On the one hand, the work is often highly scripted to achieve policy objec-
tives that have their origins in the political process. On the other hand, the
work requires improvisation and responsiveness to the individual case. Not
only that, but generally the public wants administrators of public services to
be at least open to the possibility that a special case is presenting itself, or
that extraordinary efforts of one sort or another are called for.
Essentially all the great reform efforts of the last thirty years to improve
performance or accountability in street-level public services may be under-
stood as attempts to manage this apparently paradoxical reality: how to treat
all citizens alike in their claims on government, and how at the same time to
be responsive to the individual case when appropriate. The phrase "street-
level bureaucracy" hints at this paradox. "Bureaucracy" implies a set of rules
and structures of authority; "street-level" implies a distance from the center
where authority presumably resides.
In Street-Level Bureaucracy, I show how people experience public poli-
cies in realms that are critical to our welfare and sense of community. Too
often we read about education, policing, social work, and other vital public
services without realizing or being given concrete understanding of how
these public policies result from the aggregation of the separate actions of

xii
 Dilemmas of the Individual in Public Services
many individuals, or how and why the actions in question are consistently
reproduced by the behavior of individuals.
The book is grounded in observations of the collective behavior of public
service organizations and advances a theory of the work of street-level bu-
reaucracies as individual workers experience it. I argue that the decisions of
street-level bureaucrats, the routines they establish, and the devices they
invent to cope with uncertainties and work pressures, effectively become the
public policies they carry out. I maintain that public policy is not best un-
derstood as made in legislatures or top-floor suites of high-ranking adminis-
trators. These decision-making arenas are important, of course, but they do
not represent the complete picture. To the mix of places where policies are
made, one must add the crowded offices and daily encounters of street-level
workers. Further, I point out that policy conflict is not only expressed as the
contention of interest groups, as we have come to expect. It is also located
in the struggles between individual workers and citizens who challenge or
submit to client-processing.
For example, many people are convinced that police officers dispropor-
tionately single out African Americans for scrutiny and wrongly use skin
color and racial characteristics to target blacks for attention. Police officials
invariably deny that they engage in racial profiling, and suggest that if blacks
are stopped disproportionately it is because they act in ways that legiti-
mately trigger police inquiry. It is evident that, to the extent racial profiling
exists, it arises not from official policy or direct racial orientations but out of
the ways police officers draw on social stereotypes in exercising the discre-
tion sanctioned by their departments.
Similarly, we know that service bureaucracies consistently favor some cli-
ents over others, despite official policies designed to treat people alike. To
understand how and why these organizations sometimes perform contrary
to their own rules and goals, we need to know how the rules are experi-
enced by workers in the organization, what latitude workers have in acting
on their preferences, and what other pressures they experience.
Few callings deserve greater respect than those involving public service.
As citizens we are grateful to those people who teach our children, protect
life and property, manage our natural resources, and help people in need to
access social services. These functions have evolved as hallmarks of inclu-
sive, prosperous societies throughout the world. Some street-level occupa-
tions are highly respected and well-paid. Others, such as some social work-
ers, have a more contested position in society. Some operate in relative
obscurity, whereas others, such as police officers and child protection work-

xiii
Preface
ers, are among those who are often in the news for regrettable develop-
ments. If a child dies while in protective care, or a person is badly treated
while in custody, everyone in those agencies experiences the resulting pub-
lic criticism.
One important way in which street-level bureaucrats experience their
work is in their struggle to make it more consistent with their strong com-
mitments to public service and the high expectations they have for their
chosen careers. People often enter public employment with a commitment
to serving the community. Teachers, social workers, public defenders, and
police officers partly seek out these occupations because they offer socially
useful roles. Yet the very nature of these occupations can prevent recruits
to street-level bureaucracies from coming even close to the ideal concep-
tion of their jobs. Large classes, huge caseloads, and other challenging
workload pressures combine with the contagious distress of clients who
have few resources and multiple problems to defeat their aspirations as ser-
vice workers.
Ideally and by training, street-level bureaucrats should respond to the in-
dividual needs or characteristics of the people they serve or confront. In
practice, they must deal with clients collectively, because work require-
ments prohibit individualized responses. Teachers should respond to the
needs of the individual child; in practice, they must develop techniques to
manage a classroom of children. Police officers should respond to the pre-
senting case; in reality, they must develop techniques to recognize and re-
spond to various types of confrontations, particularly those that threaten
their authority or may pose danger. At best, street-level bureaucrats invent
modes of mass processing that more or less permit them to deal with the
public fairly, appropriately, and thoughtfully. At worst, they give in to favor-
itism, stereotyping, convenience, and routinizing-all of which serve their
own or agency purposes.
Compromises in work practices and attitudes are often rationalized as re-
flecting workers' greater experience on the job, their appreciation of practi-
cal and political realities, or their more realistic assessment of the nature of
the work. But these rationalizations only summarize the prevailing struc-
tural constraints on human service bureaucracies. They are not "true" in an
absolute sense. The teacher who psychologically abandons her commitment
to help every child to read may succumb to a private assessment of the sta-
tus quo in education. But this compromise says nothing about the potential
of individual children to learn or the capacity of the teacher to instruct. This
potential remains intact. It is the system of schooling, the organization of
the schooling bureaucracy, that teaches that children are developmentally

xiv
 Dilemmas of the Individual in Public Services
'"slow'" or unmotivated, and that teachers must abandon their early com-
mitments to be an excellent teacher to every child.
In the same way, the criminal justice system allows police recruits to pre-
sume that they can approach with impunity young people hanging out in
certain neighborhoods to see whether they are in possession of guns or
drugs, even if they have no evident cause for suspicion other than the coin-
cidence of age, race, and neighborhood. Young police officers learn that
judges will back them up if the young people claim that the officers planted
evidence or made up their own descriptions of the encounters. Court offi-
cers, judges, prosecutors, and public defenders collaborate in the mass pro-
cessing of a great many new and repeat juvenile offenders each year yet re-
tain the ideal that each may have his or her fair and full "day in court."
Some street-level bureaucrats drop out or bum out relatively early in
their careers. Those who stay on, to be sure, often grow in the jobs and per-
fect treatment and client-processing techniques that provide an acceptable
balance between public aspirations for the work and the coping require-
ments of the job. These adjustments of work habits and attitudes may re-
flect lower expectations for themselves, their clients, and the potential of
public policy. Ultimately, these adjustments permit acceptance of the view
that clients receive the best that can be provided under prevailing circum-
stances.
Street-level bureaucrats often spend their work lives in these corrupted
worlds of service. They believe themselves to be doing the best they can
under adverse circumstances, and they develop techniques to salvage ser-
vice and decision-making values within the limits imposed on them by the
structure of the work. They develop conceptions of their work and of their
clients that narrow the gap between their personal and work limitations and
the service ideal. These work practices and orientations are maintained
even as they contribute to the distortion of the service ideal or put the
worker in the position of manipulating citizens on behalf of the agencies
from which citizens seek help or expect fair treatment.
Should teachers, police officers, or social workers look for other work
rather than participate in practices that seem far from ideal? This would
mean leaving clients to others who have even fewer concerns and less inter-
est in clients than they do. It would mean not only starting over in a new
career, but also abandoning the satisfactory aspects of the work they have
managed to carve out.
Should they stay in their jobs and dedicate themselves to changing client-
processing conditions from within their agencies? This approach is prob-
lematic as well, though it is the career path taken by many who leave direct

XV
Preface
service for management. In their new positions, some will be reformers
striving for change to the limit of their capacity and what the times will
bear. Others will settle for the status quo.
The structure of street-level bureaucracy also confronts clients with di-
lemmas bearing on action. Consumers of public services, for the most part,
cannot choose the public services to which they will be subject. They must
accept the schools, courts, and police forces of their communities. If they
are poor, they must also accept the community's arrangements for health
care, income support, housing subsidies and other benefit programs. In ap-
proaching the institutions that administer these policies, they must strike a
balance between asserting their rights as citizens and conforming to the be-
haviors public agencies seek to place on them as clients. As citizens, they
should seek all to which they are entitled. As bureaucratic subjects, they
must temper their demands in accord with their assessment of the limita-
tions of the public agencies which control benefits and sanctions. Although
it is apparent that exceptions are often made and additional resources often
found, clients also recognize the potential costs of unsuccessfully asserting
their rights.
On matters of the greatest urgency and moment, such as health care, ed-
ucation, justice, housing, and income, clients passively seek support and fair
treatment from public agencies when evidence and experience suggest that
their hopes may go unrewarded. The dilemmas of action may be particu-
larly acute if clients are poor, are immigrants, or are of a different racial or
ethnic background than the public employees with whom they interact.
Should I wait my tum and submit to the procedures of the agency, despite
reservations? I risk being unable to gain attention to my particular needs
and concerns. Should I speak out forcefully and demand my rights? I risk
antagonizing the workers by disrupting office procedures.
It is no small thing to adjust successfully to the rigors of the street-level
workplace. Virtually all jobs involve adjustments to routines of practice,
challenges to keeping a fresh outlook despite repetitive tasks, and compro-
mises between personal needs and vocational requirements. Despite the
many barriers to effective practice described in these pages, street-level bu-
reaucrats frequently manage to find a satisfactory balance between the re-
alities of the job and personal fulfillment. The society is all the better for
their capacity to find a satisfactory balance in their work life.
When I originally wrote this book, I was intent on elaborating on the cop-
ing behaviors of street-level bureaucrats. In doing so I emphasized the gap
between the realities of practice and service ideals. This approach had its
value, to judge from the reception the book has enjoyed over the years. But

xvi
 Dilemmas of the Individual in Public Services
it led to neglect of an important reality: vast numbers of people in public
service on a daily basis go to work at rewarding and fulfilling jobs. They
meet their classes, carry out their assignments, and manage their caseloads
without much complaint. Work goes on in these public service organizations
to general satisfaction. Partly this is because workers, clients, and the gen-
eral public have more modest expectations than they might have. But it also
goes on because frontline workers have figured out how to do what they re-
gard as a reasonable job with resources at their disposal.
How does one reconcile a clear-eyed assessment of the dilemmas of
street-level bureaucracy with the reality that many if not most teachers, po-
lice officers, and social workers like what they do and do their jobs relatively
well by community standards?
There are two ways to understand the term "street-level bureaucrat."
One is to equate it with the public services with which citizens typically in-
teract. In this sense, all teachers, police officers, and social workers in pub-
lic agencies are street-level bureaucrats without further qualification. This is
the way the term has commonly come to be used.
Another way-the one I originally intended-was to define street-level
bureaucracy as public service employment of a certain sort, performed un-
der certain conditions. In this second approach, street-level bureaucrats in-
teract with citizens in the course of the job and have discretion in exercising
authority; in addition, they cannot do the job according to ideal conceptions
of the practice because of the limitations of the work structure.
When I first wrote the book, I did not mean to suggest that every front-
line worker experienced stressful working conditions. I know teachers who
with little effort are able to pay attention to every child in their small school.
I know art teachers who experience little stress in leading their classrooms
through exercises. I know National Park rangers whose daily routines re-
quire coping with boredom as much as anything else.
In this second conception of the term, in other words, not every teacher,
police officer, or public social worker experiences the pressures that I stated
street-level bureaucrats face by definition. Frontline workers whose jobs are
relatively free of restrictive structural constraints will still develop routines
in response to their work requirements. But the routines will not be devel-
oped primarily to cope with a difficult work environment. If we adopt the
second perspective, we can see that not every frontline worker experiences
the pressures this book analyzes.
Additionally, although many street-level coping behaviors may widen the
gap between policy as written and policy as performed, other coping behav-
iors reflect acceptable compromises between the goals of enacted policy

xvii
Preface
and the needs of the street-level workers. Not every coping mechanism dis-
tances the worker from the goals of the organization. Indeed, the best work-
ers are the ones who bridge the gap.
Perhaps it is best to imagine a continuum of work experiences ranging
from those that are deeply stressful and the processing of clients is severely
underresourced, to those that provide a reasonable balance between job re-
quirements and successful practice. Workers' places on that continuum
might change over time as they gain experience, as caseloads and assign-
ments vary, or as the workplace itself adopts new approaches or engages
new clienteles. All street-level bureaucrats potentially confront circum-
stances that lead them to coping mechanisms representing departures from
the service ideal. But all frontline workers do not cope with these issues all
the time.
Still another reason that many street-level bureaucrats can successfully
negotiate the gap between the ways they cope with their jobs and public
expectations is that those expectations are undoubtedly considerably lower
than the ideal. Public expectations may replicate on a societal basis the
compromises street-level bureaucrats adopt in coping with their clients per-
son by person.
This can be explained partly by the fact that the work of street-level bu-
reaucrats is mostly hidden from public view, so even attentive citizens do
not necessarily know what is going on agency by agency. Also, to the extent
that general expectations of public services go beyond demands for effi-
ciency and honest administration, they are likely to focus on incremental
improvement. That is, the hopes of the public for improved agency perfor-
mance are likely to focus on marginal changes in client or administrative
outcomes, and are likely to be based on limited indicators. Reformers who
hold out for prospects of radically better services and client outcomes tend
to be dismissed as excessively idealistic.
A final set of dilemmas confronts citizens who are continuously, if implic-
itly, asked to evaluate public services. This occurs in focused forums such as
a referendum on a school budget or a revolt against high property taxes. It
also occurs in diffuse expressions of dissatisfaction with the public sector,
such as Colorado's famous Taxpayers' Bill of Rights (TABOR), which set in
motion a drastic decline in public services until TABOR was suspended by
voters in 2005. Indeed, the many initiatives to limit state and local spending
in recent decades have largely been understood as attacks on the value of
government.
What are the policy alternatives? When all the "fat" has been trimmed
from agency budgets and all the "waste" eliminated, the basic choices re-

xviii
 Dilemmas of the Individual in Public Services
main: to further automate, systematize, and regulate the interactions be-
tween government employees and citizens seeking help; to drift with the
current turmoil that favors reduced services and greater standardization in
the name of cost effectiveness and budgetary controls; or to secure or re-
store the importance of human interactions in services that require discre-
tionary intervention or involvement.
How much can human intervention be eliminated from teaching, nurs-
ing, policing, and judging? The fact is that we must have people making
decisions and treating other citizens in the public services. We are not pre-
pared as a society to abandon decisions about people and discretionary in-
tervention to machines and programmed formats. Yet how can one advocate
greater attention to the intervening and discretionary roles of street-level
bureaucrats in the face of the enormous and often well-deserved popular
discontent with the effectiveness and quality of their work?
I try to address these questions in this book. I do not exonerate street-
level bureaucracies, excuse their deficiencies, or urge their support as cur-
rently structured. Rather, I locate the problem of street-level bureaucrats in
the structure of their work, and attempt to identifY conditions that would
better support a reconstituted public sector dedicated to appropriate ser-
vice and respect for clients--one that would be more likely to produce ef-
fective service providers. In developing the street-level bureaucracy frame-
work, I identifY the common elements of occupations as apparently
disparate as, say, police officer and social worker. The analysis of street-level
bureaucracy helps us identifY which features of people-processing are com-
mon, and which are unique, to the different occupational milieux in which
they arise.
Moreover, this essentially comparative approach permits us to raise ques-
tions systematically about apparent differences in various service areas. For
example, recognition that all street-level bureaucracies need to control cli-
ents gives perspective to police officer shows of force and raises questions
about precisely what in the work context of police officers makes client con-
trol so dominant a theme.
Just as one of the most important contributions of the concept of "profes-
sionalism" is to facilitate understanding of the differences between, say,
doctors and nurses, in the same way the concept of street-level bureaucracy
should encourage exploration of important differences in public services as
well as contribute to an understanding of central tendencies that they share.
Street-level bureaucrats are major recipients of public expenditures and
represent a significant portion of public activity at the local level. Citizens
directly experience government through them, and their actions are the

xix
Preface
policies provided by government in important respects. I start by summariz-
ing the importance of street-level bureaucrats in contemporary political life
and explain the sense in which these low-level workers may be understood
to "make" the policies they are otherwise charged with implementing (part
I). Then I treat the common features of street-level work and explore the
implications of these conclusions for client outcomes, organizational con-
trol, and worker satisfaction (part II).
The utility of the street-level bureaucracy approach can be tested only in
efforts to understand whether common features of the framework lead to
common behavioral outcomes. I explore this general question with refer-
ence to street-level tendencies to ration and restrict services, control clients
and the work situation, and develop psychological dispositions that reduce
the dissonance between worker expectations and actual service outcomes
(part III). In the next section, I provide an assessment of the effect of fiscal
crisis on street-level bureaucrats, and a discussion of the potential for re-
form and reconstruction of these critical public functions (part IV).
These latter chapters may be of particular interest to readers of this new
edition for the insight they may provide on developments over the last thirty
years. On the one hand, the implications that reform movements within the
professions might play a restorative role today seem more farfetched than
they did thirty years ago. On the other hand, the choices available to the
society for managing street-level bureaucracies toward greater responsive-
ness and democratic accountability remain reasonably intact. It is also note-
worthy that the theme of fiscal crisis, which dominated discussions of cut-
backs in public services as a result of tax revolts of the late 1970S, are still
with us today. These themes are reviewed and account taken of recent de-
velopments in public services in the final chapter, which was written espe-
cially for this edition.

XX
 ACKNOWLEDGMENTS

2010

For more than thirty years I have been privileged to track the development
of street-level bureaucracy as the ideas embodied in this book have rolled
out, been used and evaluated, and deployed in new ways. They serve some
simply to designate classes of actors in the policy process. In other hands
they have been adopted and the framework modified to guide extensive in-
quiries into very specific lines of work. Researchers have organized their re-
search based on the street-level bureaucracy perspective in the study of cus-
toms inspectors in Senegal, employment counselors in Australia, and labor
inspectors in the Dominican Republic. It has been extremely gratifying to
have been able to observe closely this swirl of ideas.
Of all the comments and critiques I have received during this long pe-
riod, two have remained particularly memorable. The first is a set of ac-
counts of many current and former public sector workers who have read the
book, usually on the occasion of having returned to graduate school after a
few years in the field. They say that in reading the book they recognize
themselves and their struggles at work. They report that the book helped
them feel better about the way they adapted to life in the organization. The
difficulties they were having at work, they now understood, were not neces-
sarily attributable to their personal failings, but instead at least in part were
the result of the structure of their jobs. I particularly appreciate these com-
ments because a first step in empowerment of the individual is recognizing
the systemic basis of one's condition or circumstances.
As to the second, a few years after the book was published I agreed to be
interviewed by telephone from my office at MIT by students in North Da-
kota who were studying to be social workers. One student thought the book
was very persuasive but, she said essentially, "You paint such a grim pic-
ture-after reading your book I don't know whether I want to go into the
field!" I was taken aback, but she was right. Whatever the value of the book

xxi
Acknowledgments
for researchers and policy analysts, I understood that for people considering
careers in public service the book might well be discouraging.
I have tried to address this concern here, in the new preface to the book,
but the topic deserves much greater attention. Literally millions of people
choose to go into the public sector because of the rewards and challenges of
working with and for other people. They deserve much more public ap-
proval and respect (and usually higher pay) than they generally receive. In
support of these choices, it would be good to know, from systematic re-
search conducted from a street-level bureaucracy perspective, how police
officers, teachers, and social workers find that satisfactory balance between
the expectations of the job and what they are able to accomplish.
I thank the Russell Sage Foundation for keeping the book in print over
all these years, and inviting me to revise the preface and write a new chap-
ter for this edition. I am very grateful to Robert Behn, Evelyn Brodkin, Car-
olyn Hill, Deborah Stone, Steven Rathgeb Smith, and Soeren Winter,
friends and distinguished scholars, who offered comments on a draft of the
new chapter on painfully short notice. I owe special thanks and recognition
to my colleagues at Demos, who provide a remarkable home for researchers
and activists committed to striving for an inclusive democracy and shared
prosperity, supported by an effective and responsive public sector.

1g8o

My interest in the common work characteristics of street-level bureau-

crats was first prompted in 1969 while writing a review of a book on the
police. That year I wrote a paper (and later published), "Toward a
Theory of Street-Level Bureaucracy," recording my initial thoughts and
speculations on the importance of work structure in establishing the rela-
tionship between citizens and these public employees.
This book presents the theory toward which I was pointing in the origi-
nal essay. The book partially reformulates and greatly expands my earlier
statement and takes up many new considerations, such as the implications
of the fiscal crisis for street-level bureaucracy, that were not contemplated
earlier.
I have been greatly aided in this work by a grant from the Russell Sage
Foundation to the M.I.T.-Harvard Joint Center for Urban Studies, and
by the hospitality at different times of the Institute for Research on Pov-

xxii
 Acknowledgments
erty at the University of Wisconsin, and the Department of Political Sci-
ence and the Graduate School of Public Affairs of the University of
Washington. Graduate students at the University of Washington, students
at the College of Public and Community Service of the University of
Massachusetts, Boston, as well as graduate students at M.I.T. have taught
me a great deal about street-level bureaucracy over this period.
Many friends and colleagues have contributed to this work in conver-
sation, in their writings, or through expressions of personal interest and
support. I am particularly grateful in various measures to Robert Alford,
Gary Bellow, Murray Edelman, Willis Hawley, Ira Katznelson, Jeanne Ket-
tleson, Margaret Levi, Hannah Lipsky, David J. Olson, Jeffrey Pressman,
Martin Rein, Charles Sabel, and Aaron Wildavsky. My debt to many
other writers who have written usefully on public services that function
as street-level bureaucracy is recognized in the notes. I am pleased to
acknowledge my double debt to Martha Wagner Weinberg, first as tire-
less and inventive research assistant, and second (many years later) as val-
ued colleague at M .I. T.
Many of the ideas that comprise this book were developed in collabora-
tion with Carl Hosticka, Jeffrey Prottas, and Richard Weatherley when
they served as research assistants under the Russell Sage grant. I am proud
of that association and deeply appreciate their many insights and contri-
butions to our common enterprise.
This book has been greatly influenced by Suzanne Lipsky. Among her
many contributions has been her recognition and analysis of the potential
of people to sustain and recover their humanity despite contributing
to or being subjects of oppressive social systems.

xxiii
 PART I

INTRODUCTION
 CHAPTER 1

The Critical Role of

Street-Level Bureaucrats

Public service workers currently occupy a critical position in American soci-

ety. Although they are normally regarded as low-level employees, the ac-
tions of most public service workers actually constitute the services "deliv-
ered" by government. Moreover, when taken together the individual
decisions of these workers become, or add up to, agency policy. Whether
government policy is to deliver "goods"-such as welfare or public
housing~r to confer status-such as "criminal" or "mentally ill"-the dis-
cretionary actions of public employees are the benefits and sanctions of gov-
ernment programs or determine access to government rights and benefits.
Most citizens encounter government (if they encounter it at all) not
through letters to congressmen or by attendance at school board meetings
but through their teachers and their children's teachers and through the po-
liceman on the comer or in the patrol car. Each encounter of this kind repre-
sents an instance of policy delivery.
Public service workers who interact directly with citizens in the course of
their jobs, and who have substantial discretion in the exe.cution of their work
are called street-level bureaucrats in this study. Public service agencies that
employ a significant number of street-level bureaucrats in proportion to
their work force are called street-level bureaucracies. Typical street-level bu-
reaucrats are teachers, police officers and other law enforcement personnel,
social workers, judges, public lawyers and other court officers, health work-
ers, and many other public employees who grant access to government pro-
grams and provide services within them. People who work in these jobs tend

3
INTRODUCTION

to have much in common because they experience analytically similar work

conditions. 1
The ways in which street-level bureaucrats deliver benefits and sanctions
structure and delimit people's lives and opportunities. These ways orient
and provide the social (and political) contexts in which people act. Thus
every extension of service benefits is accompanied by an extension of state
influence and control. As providers of public benefits and keepers of public
order, street-level bureaucrats are the focus of political controversy. They
are constantly tom by the demands of service recipients to improve effec-
tiveness and responsiveness and by the demands of citizen groups to im-
prove the efficacy and efficiency of government services. Since the salaries of
street-level bureaucrats comprise a significant proportion of nondefense gov-
ernmental expenditures, any doubts about the size of government budgets
quickly translate into concerns for the scope and content of these public ser-
vices. Moreover, public service workers have expanded and increasingly
consolidated their collective strength so that in disputes over the scope of
public services they have become a substantial independent force in the res-
olution of controversy affecting their status and position.
Street-level bureaucrats dominate political controversies over public ser-
vices for two general reasons. First, debates about the proper scope and
focus of governmental services are essentially debates over the scope and
function of these public employees. Second, street-level bureaucrats have
considerable impact on peoples' lives. This impact may be of several kinds.
They socialize citizens to expectations of government services and a place in
the political community. They determine the eligibility of citizens for gov-
ernment benefits and sanctions. They oversee the treatment (the service)-
citizens receive in those programs. Thus, in a sense street-level bureaucrats
implicitly mediate aspects of the constitutional relationship of citizens to the
state. In short, they hold the keys to a dimension of citizenship.

C onfiict over the Scope and Substance of Public Services

In the world of experience we perceive teachers, welfare workers, and police

officers as members of separately organized and motivated public agencies.
And so they are from many points of view. But if we divide public employees
according to whether they interact with citizens directly and have discretion
over significant aspects of citizens' lives, we see that a high proportion and

4
 The Critical Role of Street-Level Bureaucrats
enormous number of public workers share these job characteristics. They
comprise a great portion of all public employees working in domestic affairs.
State and local governments employ approximately 37 million people in
local schools, more than 500,000 people in police operations, and over
300,000 people in public welfare. Public school employees represent more
than half of all workers employed in local governments. Instructional jobs
represent about two-thirds of the educational personnel, and many of the
rest are former teachers engaged in administration, or social workers, psy-
chologists, and librarians who provide direct services in the schools. Of the
3.2 million local government public employees not engaged in education,
approximately 14 percent work as police officers. One of every sixteen jobs
in state and local government outside of education is held by a public welfare
worker. 2 In this and other areas the majority of jobs are held by people with
responsibility for involvement with citizens.
Other street-level bureaucrats comprise an important part of the re-
mainder of local government personnel rolls. Although the U.S. Census
Bureau does not provide breakdowns of other job classifications suitable for
our purposes, we can assume that many of the 1.1 million health workers, 3
most of the 5,000 public service lawyers, 4 many of the employees of the
various court systems, and other public employees also perform as street-
level bureaucrats. Some of the nation's larger cities employ a staggering
number of street-level bureaucrats. For example, the 26,68o school teachers
in Chicago are more numerous than the populations of many of the Chicago
suburbs. 5
Another measure of the significance of street-level bureaucrats in public
sector employment'is the amount of public funds allocated to pay them. Of
all local government salaries, more than half went to public education in
1973. Almost 8o percent of these monies was used to pay instructional per-
sonnel. Police salaries comprised approximately one-sixth of local public
salaries not assigned to education. 6
Much of the growth in public employment in the past 25 years has oc-
curred in the ranks of street-level bureaucrats. From 1955 to 1975 govern-
ment employment more than doubled, largely because the baby boom of the
postwar years and the growing number of elderly, dependent citizens in-
creased state and local activity in education, health, and public welfare. 7
Street-level bureaucracies are labor-intensive in the extreme. Their busi-
ness is providing service through people, and the operating costs of such
agencies reflect their dependence upon salaried workers. Thus most of what-
ever is spent by government on education, police, or other social services
(aside, of course, from income maintenance, or in the case of jails and

5
INTRODUCTION

prisons, inmate upkeep) goes directly to pay street-level bureaucrats. For

example, in large cities over go percent of police expenditures is used to pay
for salaries. 8
Not only do the salaries of street-level bureaucrats constitute a major por-
tion of the cost of public services, but also the scope of public services
employing street-level bureaucrats has increased over time. Charity was
once the responsibility of private agencies. The federal government now
provides for the income needs of the poor. The public sector has absorbed
responsibilities previously discharged by private organizations in such di-
verse and critical areas as policing, education, and health. Moreover, in all
these fields government not only has supplanted private organizations but
also has expanded the scope of responsibility of public ones. This is evident
in increased public expectations for security and public safety, the extension
of responsibilities in the schools to concerns with infant as well as 'post-
adolescent development, and public demands for affordable health care
services. 9
Public safety, public health, and public education may still be elusive
social objectives, but in the past century they have been transformed into
areas for which there is active governmental responsibility. The transforma-
tion of public responsibility in the area of social welfare has led some to rec.
ognize that what people "have" in modern American society often may con-
sist primarily of their claims on government "largesse," and that claims to
this "new property" should be protected as a right of citizens. 10 Street-level
bureaucrats play a critical role in these citizen entitlements. Either they di-
rectly provide public benefits through services, or they mediate between cit-
izens and their new but by no means secure estates.
The poorer people are, the greater the influence street-level bureaucrats
tend to have over them. Indeed, these public workers are so situated that
they may well be taken to be part of the problem of being poor. Consider the
welfare recipient who lives in public housing and seeks the assistance of a
legal services lawyer in order to reinstate her son in school. He has been sus-
pended because of frequent encounters with the police. She is caught in a
net of street-level bureaucrats with conflicting orientations toward her, all
acting in what they call her "interest" and "the public interest." 11
People who are not able to purchase services in the private sector must
seek them from government if they are to receive them at all. Indeed, it is
taken as a sign of social progress that poor people are granted access to ser-
vices if they are too poor to pay for them.
Thus, when social reformers seek to ameliorate the problems of the poor,
they often end up discussing the status of street-level bureaucrats. Welfare

6
 The Critical Role of Street-Level Bureaucrats
reformers move to separate service provision from decisions about support
payments, or they design a negative income tax system that would eliminate
social workers in allocating welfare. Problems of backlog in the courts are
met with proposals to increase the number of judges. Recognition that early-
childhood development largely establishes the potential for later achieve-
ment results in the development of new programs (such as Head Start) in
and out of established institutions, to provide enriched early-childhood
experiences.
In the 1g6os and early 1970s the modal governmental response to social
problems was to commission a corps of street-level bureaucrats to attend to
them. Are poor people deprived of equal access to the courts? Provide them
with lawyers. Equal access to health care? Establish neighborhood clinics.
Educational opportunity? Develop preschool enrichment programs. It is far
easier and less disruptive to develop employment for street-level bureau-
crats than to reduce income inequalities.
In recent years public employees have benefitted considerably from the
growth of public spending on street-leve! bureaucracies. 12 Salaries have
increased from inadequate to respectable and even desirable. Meanwhile,
public employees, with street-level bureaucrats in the lead, have secured
unprecedented control over their work environments through the develop-
ment of unions and union-like associations. 13 For example, teachers and
other instructional personnel have often been able to maintain their posi-
tions and even increase in number, although schools are more frequently
under attack for their cost to taxpayers. The ratio of instructional personnel
in schools has continued to rise despite the decline in the number of school-
age children. 14 This development supplements general public support for
the view that some street-level bureaucrats, such as teachers and police of-
ficers, are necessary for a healthy society. us
The fiscal crisis that has affected many cities, notably New York and more
recently Cleveland and Newark, has provided an opportunity to assess the
capacity of public service workers to hold onto their jobs in the face of enor-
mous pressures. Since so much of municipal budgets consists of inflexible,
mandated costs-for debt service., pension plans and other personnel bene-
fits, contractually obligated salary increases, capital expenditure commit-
ments, energy purchases, and so on-the place to find "fat" to eliminate
from municipal budgets is in the service sector, where most expenditures
tend to be for salaries. While many public employees have been fired during
this crisis period, it is significant that public service workers often have been
able to lobby, bargain, and cajole to minimize this attrition. 18 They are sup-
ported in their claims by a public fearful of a reduced police force on the

7
INTRODUCTION

street and resentful of dirtier streets resulting from fewer garbage pickups.
They are supported by families whose children will receive less instruction
from fewer specialists than in the past if teachers are fired. And it does not
hurt their arguments that many public employees and their relatives vote in
the city considering force reductions. 17
The growth of the service sector represents the furthest reaches of the
welfare state. The service sector penetrates every area of human needs as
they are recognized and defined, and it grows within each recognized area.
This is not to say that the need is met, but only that the service state
breaches the barriers between public responsibility and private affairs.
The fiscal crisis of the cities focuses on the service sector, fundamentally
challenging the priorities of the service state under current perceptions of
scarcity. Liberals have now joined fiscal conservatives in challenging service
provision. They do not do so directly, by questioning whether public ser-
vices and responsibilities developed in this century are appropriate. Instead,
they do it backhandedly, arguing that the accretion of public employees and
their apparently irreversible demands upon revenues threaten the au-
tonomy, flexibility, and prosperity of the political order. Debates over the
proper scope of services face the threat of being overwhelmed by challenges
to the entire social service structure as seen from the perspective of unbal-
anced public budgets.

Conflict over Interactions with Citizens

I have argued that street-level bureaucrats engender controversy because

they must be dealt with if policy is to change. A second reason street-level
bureaucrats tend to be the focus of public controversy is the immediacy of
their interactions with citizens and their impact on peoples' lives. The policy
delivered by street-level bureaucrats is most often immediate and personal.
They usually make decisions on the spot (although sometimes they try not to)
and their determinations are focused entirely on the individual. In contrast,
an urban renewal program might destroy a neighborhood and replace and
substitute new housing and different people, but the policy was prolonged,
had many different stages, and was usually played out in arenas far removed
from the daily life of neighborhood residents.
The decisions of street-level bureaucrats tend to be redistributive as well
as allocative. By determining eligibility for benefits they enhance the claims

8
 The Critical Role of Street-Lev el Bureaucra ts
of some citizens to government al goods and services at the expense of gen-
eral taxpayers and those whose claims are denied. By increasing or decreas-
ing benefits availability to low-income recipient populations they implicitly
regulate the degree of redistributio n that will be paid for by more affluent
sectors.
In another sense, in delivering policy street-level bureaucrats make deci-
sions about people that affect their life chances. To designate or treat some-
one as a welfare recipient, a juvenile delinquent, or a high achiever affects
the relationships of others to that person and also affects the person's self-
evaluation. Thus begins (or continues) the social process that we infer
accounts for so many self-fulfilling prophecies. The child judged to be ajuve-
nile delinquent develops such a self-image and is grouped with other "delin-
quents," increasing the chances that he or she will adopt the behavior
thought to have been incipient in the first place. Children thought by their
teacher to be richly endowed in learning ability learn more than peers of
equal intelligence who were not thought to be superior. 18 Welfare recipients
find or accept housing inferior to those with equal disposable incomes who
are not recipients.19
A defining facet of the working environmen t of street-level bureaucrats is
that they must deal with clients' personal reactions to their decisions, how-
ever they cope with their implications. To say that people's self-evaluation is
affected by the actions of street-level bureaucrats is to say that people are
reactive to the policy. This is not exclusively confined to subconscious pro-
cesses. Clients of street-level bureaucracie s respond angrily to real or per-
ceived injustices, develop strategies to ingratiate themselves with workers,
act grateful and elated or sullen and passive in reaction to street-level bu-
reaucrats' decisions. It is one thing to be treated neglectfully and routinely
by the telephone company, the motor vehicle bureau, or other government
agencies whose agents know nothing of the personal circumstanc es sur-
rounding a claim or request. It is quite another thing to be shuffied, catego-
rized, and treated "bureaucrati caily," (in the pejorative sense), by someone
to whom one is directly talking and from whom one expects at least an open
and sympathetic hearing. In short, the reality of the work of street-level bu-
reaucrats could hardly be farther from the bureaucratic ideal of impersonal
detachment in decision making. 20 On the contrary, in street-level bureau-
cracies the objects of critical decisions-p eople-actua lly change as a result
of the decisions.
Street-level bureaucrats are also the focus of citizen reactions because
their discretion opens up the possibility that they will respond favorably on
behalf of people. Their general and diffuse obligation to the "public interest"

9
INTRODUCT ION

pennits hope to flourish that the individual worker will adopt a benign or fa-
vorable orientation toward the client. Thus, in a world of large and imper-
sonal agencies that apparently hold the keys to important benefits, sanctions,
and opportunities, the ambiguity of work definitions sustains hope for a
friend in court.
This discussion helps explain continued controversy over street-level bu-
reaucracies at the level of individual service provision. At the same time, the
peculiar nature of government service delivery through street-level bureau-
crats helps explain why street-level bureaucracies are apparently the pri-
mary focus of community conflict in the current period, and why they are
likely to remain the focus of such conflict in the foreseeable future. It is no
accident that the most heated community conflicts since 1964 have focused
on schools and police departments, and on the responsiveness of health and
welfare agencies and institutions. 21 These are the sites of the provision of
public benefits and sanctions. They are the locus of individual decisions
about and treatment of citizens, and thus are primary targets of protest. As
Frances Fox Piven and Richard Cloward explain:
... people experience deprivation and oppression within a concrete setting, not as
the end product oflarge and abstract processes, and it is the concrete experience that
molds their discontent into specific grievances against specific targets .... People on
relief [for example] experience the shabby waiting rooms, the overseer or the case-
worker, and the dole. They do not experience American social welfare policy .... In
other words, it is the daily experience of people that shapes their grievances, es-
tablishes the measure of their demands, and points out the targets of their anger. 22

While people may experience these bureaucracies as individuals, schools,

precinct houses, or neighborhood clinics are places where policy about indi-
viduals is organized collectively. These administrative arrangements suggest
to citizens the possibility that controlling, or at least affecting, their struc-
tures will influence the quality of individual treatment. Thus we have two
preconditions for successful community organization efforts: the hope and
plausibility that individual benefits may accrue to those taking part in group
action and a visible, accessible, and blamable collective target. 23
Community action focused on street-level bureaucracies is also apparently
motivated by concerns for community cl}aracter. The dominant institutions
in communities help shape community identity. They may be responsive to
the dominant community group (this has been the traditional role of high
schools in Boston) or they may be unresponsive and opposed to conceptions
of community and identity favored by residents, as in the case of schools that
neglect the Spanish heritage of a significant minority. Whether people are
motivated by specific grievances or more diffuse concerns that become

10
 The Critical Role of Street-L evel Bureaucr ats
directed at communit y institutions, their focus in protesting the actions of
street-leve l bureaucrac ies may be attributed to the familiarity of the agency,
its critical role in communit y welfare, and a perception at some level that
these institutions are not sufficiently accountable to the people they serve.
Finally, street-leve l bureaucra ts play a critical role in regulating the de-
gree of contempor ary conflict by virtue of their role as agents of social con-
trol. Citizens who receive public benefits interact with public agents who
require certain behaviors of them. They must anticipate the requireme nts of
these public agents and claimants must tailor their actions and develop "suit-
able" attitudes both toward the services they receive and toward the street-
level bureaucra ts themselve s. Teachers convey and enforce expectations of
proper attitudes toward schooling, self, and efficacy in other interactions.
Policemen convey expectations about public behavior and authority. Social
workers convey expectations about public benefits and the status of
recipients.
The social control function of street-leve l bureaucra ts requires comment
in a discussion of the place of public service workers in the larger society.
The public service sector plays a critical part in softening the impact of the
economic system on those who are not its primary beneficiaries and inducing
people to accept the neglect or inadequacy of primary economic and social
institutions. Police, courts, and prisons obviously play such a role in proces-
sing the junkies, petty thieves, muggers, and others whose behavior toward
society is associated with their economic position. It is a role equally played
by schools in socializing the population to the economic order and the likely
opportunit ies for different strata of the population. Public support and em-
ployment programs expand to ameliorate the impact of unemploy ment or
reduce the incidence of discontent ; they contract when employme nt oppor-
tunities improve. Moreover, they are designed and implemen ted to convey
the message that welfare status is to be avoided and that work, however
poorly rewarded, is preferable to public assistance. One can also see the two
edges of public policy in the "war on poverty" where the public benefits of
social service and communit y action invested neighborh ood institutions with
benefits for which potential dissidents could compete and ordinary citizens
could develop dependenc y. 24
What to some are the highest reaches of the welfare state are to others the
furthest extension of social control. Street-lev elbureauc rats are partly the
focus of controvers y because they play this dual role. Welfare reform
founders on disagreem ents over whether to eliminate close scrutiny of wel-
fare applications in order to reduce administra tive costs and harassmen t of
recipients, or to increase the scrutiny in the name of controlling abuses and
11
 INTROD UCTION

preventing welfare recipients from taking advantage. Juvenile corrections

and mental health policy founder on disputes over the desirability of disman-
tling large institutions in the name of cost effectiveness and rehabilitation, or
retaining close supervision in an effort to avoid the costs of letting unrecon-
structed "deviants" loose. In short, street-level bureaucrats are also at the
center of controversy because a divided public perceives that social control
in the name of public order and acceptance of the status quo are social objec-
tives with which proposals to reduce the role of street-level bureaucrats
(eliminating welfare checkups, reducing parole personnel, decriminalizing
marijuana) would interfere.
Public controversy also focuses on the proper kind of social control. Cur-
rent debates in corrections policy, concerning automatic sentencing and a
"hard-nosed " view of punishment or more rehabilitative orientations, reflect
conflict over the degree of harshness in managing prison populations. In ed-
ucational practice the public is also divided as to the advisability of liberal
disciplinary policies and more flexible instruction or punitive discipline and
more rigid, traditional approaches. The "medicalization" of deviance, in
which disruptive behavior is presumed cause for intervention by a doctor
rather than a disciplinarian, is another area in which there is controversy
over the appropriate kind of social control.
From the citizen's viewpoint, the roles of street-level bureaucrats are as
extensive as the functions of government and intensively experienced as daily
routines require them to interact with the street ministers of education, dis-
pute settlement, and health services. Collectively, street-level bureaucrats
absorb a high share of public resources and become the focus of society's
hopes for a healthy balance between provision of public services and a rea-
sonable burden of public expenditures . As individuals, street-level bureau-
crats represent the hopes of citizens for fair and effective treatment by gov-
ernment even as they are positioned to see clearly the limitations on
effective intervention and the constraints on responsiveness engendered by
mass processing.

12
 CHAPTER 2

Street-Level Bureaucrats
as Policy Makers

Street-level bureaucrats make policy in two related respects. They exercise

wide discretion in decisions about citizens with whom they interact. Then,
when taken in concert, their individual actions add up to agency behavior.
The task in this chapter is to demonstrate that the position of street-level bu-
reaucrats regularly permits them to make policy with respect to significant
aspects of their interactions with citizens. Later chapters will explore the
implications of making policy at the street level.
The policy-making roles of street-level bureaucrats are built upon two in-
terrelated facets of their positions: relatively high degrees of discretion and
relative autonomy from organizational authority.

Discretion

Unlike lower-level workers in most organizations, street-level bureaucrats

have considerable discretion in determining the nature, amount, and quality of
benefits and sanctions provided by their agencies. 1 Policemen decide who to
arrest and whose behavior to overlook. Judges decide who shall receive a
suspended sentence and who shall receive maximum punishment. Teachers
decide who will be suspended and who will remain in school, and they make
subtle determinations of who is teachable. Perhaps the most highly refined
example of street-level bureaucratic discretion comes from the field of cor-

13
INTRODUCTION

rections. Prison guards conventionally file injurious reports on inmates

whom they judge to be guilty of" silent insolence." Clearly what does or does
not constitute a dirty look is a matter of some subjectivity. 2
This is not to say that street-level workers are unrestrained by rules, regu-
lations, and directives from above, or by the norms and practices of their oc-
cupational group. On the contrary, the major dimensions of public policy-
levels of benefits, categories of eligibility, nature of rules, regulations and
services-are shaped by policy elites and political and administrative of-
ficials. Administrators and occupational and community norms also structure
policy choices of street-level bureaucrats. These influences establish the
major dimensions of street-level policy and account for the degree of stan-
dardization that exists in public programs from place to place as well as in
local programs.
To the extent that street-level bureaucrats are professionals, the assertion
that they exercise considerable discretimi is fairly obvious. Professionals are
expected to exercise discretionary judgment in their field. They are regu-
larly deferred to in their specialized areas of work and are relatively free
from supervision by superiors or scrutiny by clients. 3 Yet even public em-
ployees who do not have claims to professional status exercise considerable
discretion. Clerks in welfare and public housing agencies, for example, may
exercise discretion in determining client access to benefits, even though
their discretion is formally circumscribed by rules and relatively close
supervision.
Rules may actually be an impediment to supervision. They may be so vo-
luminous and contradictory that they can only be enforced or invoked selec-
tively. In most public welfare departments, regulations are encyclopedic,
yet at the same time, they are constantly being changed. With such rules ad-
herence to anything but the most basic and fundamental precepts of eligibil-
ity cannot be expected. Police behavior is so highly specified by statutes and
regulations that policemen are expected to invoke the law selectively. They
could not possibly make arrests for all the infractions they observe during
their working day. 4 (Like doctors and clergymen in many jurisdictions, they
are required to be on-duty and ready to intervene even during their off-duty
hours.) Similarly, federal civil-rights compliance officers have so many man-
dated responsibilities in comparison to their resources that they have been
free to determine their own priorities. 5 It would seem that the proliferation
of rules and responsibilities is only problematically related to the degree of
discretion street-level bureaucrats enjoy. 6
Although the case for the pervasive existence of discretion in street-level
work is fairly easy to make it is important to remember that in public services
 Street-Level Bureaucrats as Policy Makers
some interactions take place with citizens that involve relatively little bu-
reaucratic discretion. Patrolmen assigned to traffic duty or gun permit appli-
cations, for example, may interact with the public but exercise little discre-
tion in performing these tasks. Discretion is a relative concept. It follows
that the greater the degree of discretion the more salient this analysis in un-
derstanding the character of workers' behavior.
Since many of the problems discussed here would theoretically disappear
if workers' discretion were eliminated, one may wonder why discretion re-
mains characteristic of their jobs. The answer is that certain characteristics of
the jobs of street-level bureaucrats make it difficult, if not impossible, to
severely reduce discretion. They involve complex tasks for which elabora-
tion of rules, guidelines, or instructions cannot circumscribe the alterna-
tives. This may be the case for one of at least two reasons.
First, street-level bureaucrats often work in situations too complicated to
reduce to programmatic formats. Policemen cannot carry around instruc-
tions on ho~ to intervene with citizens, particularly in potentially hostile en-
counters. Indeed, they would probably not go out on the street if such in-
structions were promulgated, or they would refuse to intervene in
potentially dangerous situations. 7 Similarly, contemporary views of educa-
tion mitigate against detailed instructions to teachers on how and what to
teach, since the philosophy prevails that to a point every child requires a
response appropriate to the specific learning context.
Second, street-level bureaucrats work in situations that often require re-
sponses to the human dimensions of situations. They have discretion be-
cause the accepted definitions of their tasks call for sensitive observation and
judgment, which are not reducible to programmed formats. It may be that
uniform sentencing would reduce inequities in the criminal justice system.
But we also want the law to be responsive to the unique circumstances of in-
dividual transgressions. 8 We want teachers to perceive the unique potential
of children. In short, to a degree the society seeks not only impartiality from
its public agencies but also compassion for special circumstances and flexibil-
ity in dealing with them. 9
A third reason discretion is not likely to be eliminated bears more on the
function of lower-level workers who interact with citizens than with the na-
ture of the tasks. Street-level discretion promotes workers' self-regard and
encourages clients to believe that workers hold the key to their well-being.
For both workers and clients, maintenance of discretion contributes to the
legitimacy of the welfare-service state, although street-level bureaucrats by
no means establish the boundaries of state intervention.
The search for the correct balance between compassion and flexibility on
INTRODUCTION

the one hand, and impartiality and rigid rule-application on the other hand
presents a dialectic of public service reform. Reformers attempt to limit
worker discretion at one time, and increase it at another. In order to make
ambulance dispatch by untrained personnel more efficient, health planners
work to rationalize emergency dispatch procedures by developing a pro-
grammed format to aid in identifying health emergencies. 10 Meanwhile,
other health planners seek to replace untrained admission clerks with health
professionals in order to insure greater sensitivity to the health problems of
prospective patients. 11 Programmed learning materials are introduced to
release teachers for more intensive work with some student&, while all stu-
dents benefit from these motivational and self-paced features. Later this in-
novation is found wanting because it eliminated teachers' feedback to stu-
dents and encouraged regimentation rather than individualized learning. 12
To the extent that tasks remain complex and human intervention is consid-
ered necessary for effective service, discretion will remain characteristic of
many public service jobs.

Relative Autonomy from Organizational Authority

Most analysts take for granted that the work of lower-level participants will
more or less conform to what is expected of them. Organizational theorists
recognize that there will always be some slippage between orders and the
carrying out of orders, but this slippage is usually attributed to poor com-
munication or workers' residual, and not terribly important, disagreement
with organizational goals. In any event, such difficulties are usually consid-
ered unimportant enough that organizations can overcome them.
This observation is partly derived from the recognition that lower level
workers' behavior in organizations, including public agencies, appears to be
cooperative. Workers for the most part accept the legitimacy of the formal
structure of authority, and they are not in a position to dissent successfully.
But what if workers do not share the objectives of their superiors? Lower-
level participants in organizations often do not share the perspectives and
preferences of their superiors and hence in some respects cannot be thought
to be working toward stated agency goals. At least this is the case when
workers are not recruited with an affinity for the organization's goals; work-
ers do not consider orders from "above" legitimate; or the incentives avail-
able to supervisors are matched by countermeasures available to lower-level

16
 Street-Level Bureaucrats as Policy Makers
participants. One can expect a distinct degree of noncompliance if lower-
level workers' interests differ from the interests of those at higher levels, and
the incentives and sanctions available to higher levels are not sufficient to
prevail. 13
Sometimes different levels of organizations are more appropriately cou-
ceived as intrinsically in conflict with each other rather than mutually re-
sponsive and supportive. 14 At times it is more useful to view lower-level
workers as having distinctly different interests and the resources to pursue
those interests. Here discrepancies between policy declarations and actual
policy would be expected and predictable. And explanations for the discrep-
ancies would be searched for not in the breakdown or inadequacy of the
compliance system but in the structure of the work situation from which
workers' "antagonistic" interests arise.
In such organizations policy may be carried out consistent with the inter-
ests of higher levels, but this should be understood as resulting from the mu-
tual adjustment of antagonistic perspectives as well as the result of shared in-
terests. In such cases the latent conflict in the interests of different levels is
suppressed or is a matter ofindifference to one or both parties. This approach
takes as problematic the mutuality of interests, and it searches instead for
the mechanisms by which essentially antagonistic or divergent interests are
adj usted. 15
Some of the ways lower-level workers can withhold cooperation within
their organizations include such personal strategies as not working (excessive
absenteeism, quitting), aggression toward the organization (stealing, cheat-
ing, deliberate wasting), and negative attitudes with implications for work
(alienation, apathy). 16 Workers may take advantage of collective resources to
act nonco.Jperatively by forming trade unioos or by exercising rights under
collective bargaining agreements or civil service regulations. These collec-
tive strategies for noncooperation contribute to workers' willingness to dis-
play lack of motivation and to perform at only minimal levels. 17
These forms of noncooperation injure organizations' abilities to achieve
their objectives because workers perform at less than full capacity. The man-
agement challenge perceived to be at the heart of the problem is how to
make workers' needs for personal, material, or psychological gratification
mesh with the organization's needs. Thus the management problem regard-
ing worker absenteeism becomes how to improve job satisfaction while re-
taining productivity.
However, there is another class of conflicts between lower-level workers
and the organizations that arise not from the personal needs of the workers
alone but also from their positions within their organizations. The role of
INTRODUCTION
street-level bureaucrats, like other roles, may be conceived as a set of ex-
pected interests 18 as well as expected behaviors. Street-level bureaucrats
may be shown to have distinctly different interests from the interests of
others in the agencies for which they work. Moreover, certain features of
their role make it possible for them to make these differences manifest. Dif-
ferences in interests, and the possibility of surfacing the differences, permit
analysis of the structural position of street-level bureaucrats from a conflict
perspective. 19
In the following brief discussion, the nature of the conflict between the
objectives and orientations of street-level bureaucrats and those in higher
authority roles is considered first. Then the capacity of street-level bureau-
crats' ability to resist organizational directives is treated. 20

Differences between Street-Level Bureaucrats and Managers

In general lower-level workers have different job priorities than managers.

At the very least, workers have an interest in minimizing the danger and dis-
comforts of the job and maximizing income and personal gratification. These
priorities are of interest to management for the most part only as they relate
to productivity and effectiveness. In street-level bureaucracies lower-level
workers are likely to have considerably more than minimal differences with
management. Earlier it was suggested that worker compliance is affected by
the extent to which managers' orders are considered legitimate. Street-level
bureaucrats may consider legitimate the right of managers to provide direc-
tives, but they may consider their managers' policy objectives illegitimate.
Teachers asked to participate in compensatory education programs in which
they do not believe, or policemen no longer able to arrest derelicts for al-
coholism, may resist these policy objectives in various ways.
One way in which the interests of street-level bureaucrats depart from
those of managers is their need to process work loads expeditiously, free
from real and psychological threats. The fact that street-level bureaucrats
must exercise discretion in processing large amounts of work with inade-
quate resources means that they must develop shortcuts and simplifications
to cope with the press of responsibilities. The coping mechanisms street-
level bureaucrats develop (see part III) often are unsanctioned by man-
agers of their agencies.
Managers are interested in achieving results consistent with agency objec-
 Street-Level Bureaucrats as Policy Makers
tives. Street-level bureaucrats are interested in processing work consistent
with their own preferences and only those agency policies so salient as to be
backed up by significant sanctions. These sanctions must be limited. If ev-
erything receives priority, nothing does. Work-processing devices are part
of the informal agency structure that may be necessary to maintain the orga-
nization, even though the procedures may be contrary to agency policy. 21
This is a neat paradox. Lower-level participants develop coping mecha-
nisms contrary to an agency's policy but actually basic to its survival. For ex-
ample, brutality is contrary to police policy, but a certain degree of looking-
the-other-way on the part of supervisors may be considered necessary to
persuade officers to risk assault. Street-level bureaucrats have a role interest
in securing the requirements of completing the job. Managers, on the other
hand, are properly result-oriented. They are concerned with performance,
the cost of securing performance, and only those aspects of process that ex-
pose them to critical scrutiny.
Another aspect of street-level bureaucrats' role interests is their desire to
maintain and expand their autonomy. Managers try to restrict workers' dis-
cretion in order to secure certain results, but street-level bureaucrats often
regard such efforts as illegitimate and to some degree resist them success-
fully. Indeed, to the extent that street-level bureaucrats (and this would
include police, teachers, social workers, and nurses, as well as doctors and
lawyers) expect themselves to make critical discretionary decisions, many of
managers' efforts to dictate service norms are regarded as illegitimate. To
the extent that this is the case we have uncovered a condition for non-
compliance of lower-level workers. This does not mean that efforts to con-
strain street-level bureaucrats are in fact illegitimate. Street-level bureau-
crats have some claims to professional status, but they also have a
bureaucratic status that requires compliance with superiors' directives. It
does mean, however, that street-level bureaucrats will perceive their inter-
ests as separate from managers' interests, and they will seek to secure these
interests. 22
Street-level bureaucrats conspicuously create capacities to act with discre-
tion and hang on to discretionary capacities they have enjoyed in the past.
The maintenance and enhancement of discretion is so important that some
detailed illustrations may be useful.
Lower-court judges have recently encouraged the development of a great
many alternatives to incarceration, in essence turning the courts into social
work referral services. In Massachusetts and elsewhere lower-court judges
can refer presumptive offenders to many social programs, the successful
completion of which will result in obviating their sentences. These include
INTRODUCTION

programs to provide first offenders with counseling, job training, and place-
ment assistance, and alcoholics, reckless drivers, and drug offenders with ap-
propriate counseling. In addition, judges have the services of psychiatrists,
social workers, probation officers, and others who might be able to provide
treatment as an alternative to imprisonment. These developments have
been conceived by humanitarian reformers who believe, along with many
judges, that prisons create more criminals than they deter by exposing peo-
ple to experienced crooks, and by pragmatists, who recognize that the courts
have become revolving doors of repeat appearances without deterrent effect.
It is conspicuous to court observers that these programs take a heavy bur-
den off the judge. The judge is now able to make what appears to be a con-
structive decision rather than simply to choose between the unattractive al-
ternatives of sending a person to jail or releasing the putative offender
without penalty. Indirect evidence that these programs fill critical institu-
tional needs is suggested by the Boston pretrial diversion programs. These
programs were utilized beyond their capacity by judges, sometimes without
regard for the extremely important initial interview or the relatively strin-
gent eligibility requirements the programs sought to impose in order to max-
imize effectiveness. Dependent upon judges for referrals and, indeed, for
their programs' existence, administrators found it difficult to refuse judges
who referred too many clients, or inappropriate clients, to them. 23
The Veterans Administration hospital system is a fascinating bureaucracy
because it employs doctors, the preeminent professionals, in highly rule-
bound organizations. The country's system of socialized medicine for the in-
digent veteran has developed an extremely complex series of rules because
of simultaneous congressional concern to provide veterans with hospital ser-
vices, maintain strict cost accounting, and (particularly in the past) not com-
pete with private medical practice. In large part because the VA system was
to provide hospital care, leaving to private physicians the business of office
consultations, the VA hospitals were prohibited from treating patients on an
outpatient basis. This conflicted with the doctors' prerogative to prescribe a
level of treatment appropriate to the problem presented. There are many
patients who need treatment, but not hospitalization, and to hospitalize such
people would be to deny others the hospital space.
However, there was an allowable exception to the rule limiting services to
hospitalized veterans. Under the "Pre-Bedcare" category (PBC) veterans
who required health services prior to their anticipated admission (for ex-
amp!e, blood tests prior to surgery) could be treated. Despite various
requirements intended to limit PBC treatment to those whose admission
was clearly expected, actual admission to the hospitals from PBC lists was

20
 Street-Level Bureaucrats as Policy Makers
traditionally very low. It seems that doctors, chafing under the restriction
that they could not treat patients according to their best estimate of need,
were treating patients as outpatients under the fiction that they were ex
pected to be admitted. This supposition is supported by observing what hap-
pened when the VA introduced an Ambulatory Health Care (AHC) policy,
permitting outpatient treatment for the first time. PBC admissions plum-
meted, while ABC admissions rose steadily. In one hospital AHC received
148 patients in the first five months of the program, while PBC dropped
from 122 to 73 patients during that period. Fifty-one of these patients trans-
ferred directly to the AHC program. 24 In retrospect it seems that doctors
were able to utilize existing bureaucratic structures to impose their views of
proper treatment on the organization, despite organizational efforts to cir-
cumscribe their discretion.
Street-level bureaucrats will also use existing regulations and administra~
tive provision to circumvent reforms which limit their discretion. In Decem-
ber 1968, in response to pressure from the Department of Housing and
Urban Development, the Boston Housing Authority (BHA) adopted new
tenant-selection guidelines designed to insure housing project racial integra-
tion. The plan utilized what was known as the "1-2-3 rule." To eliminate per-
sonnel discretion in assignments the 1-2-3 rule provided that prospective
tenants would be offered places only in the housing projects with the three
highest vacancy ratios. If these offers were refused, the application would be
returned to the bottom of the waiting list.
The BHA integration plan did not work. Many housing authority em-
ployees objected to assigning people to projects in which they did not want
to live. They were particularly concerned for their traditionally favored cli-
entele, the elderly, poor whites who populated the "better" BHA projects.
Among the reasons the reform did not work were that housing authority per-
sonnel were so inundated with work that proper adminstrative controls were
not feasible, and in the chaos of processing applications, those who wished to
favor some prospective tenants over others were able to do so. Housing of-
ficials took advantage of provisions for exceptions to the 1-2-3 rule, interpret-
ing reasonable provisions for flexibility in extremely liberal ways when they
wanted to. They volunteered information to favored prospective tenants
concerning how to have their applications treated as emergencies or other
high-priority categories, while routinely processing the applications of oth-
ers. Applications were frequently lost or misplaced so that workers could
favor tenants simply by locating their files and acting on them, while other
files remained unavailable for processing. Meanwhile, public housing man-
agers contributed to the sustained biases of the agency by failing to report

21
INTRODU CTION

vacancies to the central office when they occurred, not informing prospec-
tive tenants when units were available, or showing tenants they wished to
discourage only unattractive or unsafe units, although others in the project
were available. Thus the press of work combined with workers' desires to
continue to serve particular clients restored the discretionary powers the
new rules were designed to eliminate. 25
The power of street-level bureaucrats to thwart reforms by manipulating
legal loopholes i!F suggested in a strategy paper prepared by New York City
officials to reduce welfare rolls. It indicates managerial awareness of the
limits of agency reform without worker cooperation. The paper rejects the
possibility of shortening intake hours too drastically, in part because of the
chance that "welfare workers would go out of the centers and take applica-
tions from people at home, on the street, etc., since application may be
made [under existing Federal law] at agency office, in own home, by tele-
phone, by mail, or in any suitable place. " 26
A dramatic example of the instinct to retain discretionary options is pro-
vided by the response to the New York State laws sponsored by then Gover-
nor Nelson Rockefeller to impose mandatory, severe jail sentences for drug
dealers, while providing relatively minor penalties for those caught with
small amounts of drugs. The rationale for the law was to provide a strong de-
terrent to the drug trafficker. Predictably, some people arrested were not
the rapacious drug dealers whom the law was designed to deter. This was the
case, for example, with drug addicts on methadone who occasionally sold a
dosage. The law presented a dilemma for court personnel, who believed that
the mandatory, minimum sentence (life imprisonment for the methadone
hustlers) was too severe for the offense. In these cases New York District At-
torney Richard Kuh began to charge the alleged offenders not with the
crimes which they had committed, but rather with crimes for which the
punishments were compatible with what he conceived to be the severity of
the offense. In this way, the district attorney attempted to provide the dis-
cretion demanded by a just court system in the face of legislation designed
precisely to eliminate this discretion. 27
Still another source of sustained differences in the interests of street-level
bureaucrats and managers is their continuous interaction with clients and
the varying degrees of complexity in this interaction. Modem bureaucracies
gain legitimacy by (often rhetorical) commitments to standards of fairness
and equity. But street-level bureaucrats are constantly confronted with the
apparent unfairness of treating people alike (just as they recognize the obvi-
ous inequities of unequal treatment). Since individuals are so much more
than their bureaucratically relevant characteristics -age, sex, place of resi-

22
 Street-Level Bureaucrats as Policy Makers
dence, income level, etc.-a failure to recognize these differences some-
times seems unfair in itself. The public housing manager believes that there
are degrees of need for public housing not circumscribed by the categories of
eligibility. The teacher recognizes that all children deserve her or his atten-
tion but thinks some require more attention than others.
Not only are the standards of fairness insufficient to dictate levels of con-
cern, but also street-level bureaucrats, like everyone else, have personal
standards of whether or not someone is deserving. The public housing man-
ager may be more sympathetic to the elderly than to other eligibles. Hous-
ing inspectors may be sympathetic to the plight of the landlord although
nothing in the formal structure of the agency encourages such a bias. 28
Under some circumstances it may be appropriate to apply standards of
service with respect to personal characteristics. Street-level bureaucrats
enjoy considerable discretion in part because society does not want compu-
terized public service and rigid application of standards at the expense of
responsiveness to the individual situation. The New York district attorney
was praised edits>rially for contriving to circumvent the mandatory sentenc-
ing requirement. The VA doctors were acting on behalf of patients and in op-
position to limitations on practice when they assigned pre-bedcare status to
patients who did not require hospitalization. Clearly discretion provides op-
portunity to intervene on behalf of clients as well as to discriminate among
them. Yet at best bureaucracies are highly ambivalent about personalistic
service delivery. At the least it is an enduring source of conflict between the
objectives of managers and workers.

Resources for Resistance

In general, lower-level workers always possess minimal resources with

which they can resist managers' orientations or achieve a modicum of re-
sponse from managers in exchange for compliance. If nothing else, since the
costs of recruiting and training a worker are rarely trivial, there is always a
degree of noncompliance that can be threatened or realized by workers. If it
were otherwise, there would be no problem of management.
Public service workers currently enjoy the benefits of collective resources
that strengthen their position considerably. Civil service provisions greatly
reduce managers' capacity to manipulate benefits and sanctions to induce
performance. Civil service provisions for advancement, introduced origi-

2J
INTRODUCTION

nally to eliminate biases in promotion, tend to be irrelevant for the skills

they purport to test, thus eliminating incentives for meritorious perfor-
mance. The costs of firing or demoting workers tend to be so great under
civil service regulations that managers often prefer to retain workers than to
endure a prolonged period of arbitration while the post in dispute remains
unfilled, or worse, remains filled by the accused incompetent. This practice
results in a mediocre standard of public service performance. Public em-
ployee unions in some cases have increased worker protection by building
additional safeguards to capricious management decisions into the collective
bargaining process.
This does not mean that managers have no control over workers. Formal
sanctions, although costly for managers to invoke, are also costly to workers,
who thus try to avoid receiving them. Managers also can manipulate discre-
tionary perquisites they control, such as recommendations for advancement
or transfer, or shift and work assignments. They also can facilitate or retard
individual workers' efforts, granting a day off here, speeding the processing
of work there, and generally making a job more or less desirable.
Lower-level workers under some conditions enjoy additional resources
stemming from their critical positions in the organization. Sociologist David
Mechanic has suggested that under the right circumstances several factors
affect the power of lower-level workers. These factors include qualities and
characteristics such as expertise, willingness to become interested and to ex-
pend effort, and personal attractiveness. They also include structural consid-
erations such as location within the organization, which affects command of
information and access and command over organizational tools. 29 These re-
sources enhance the power oflower-status workers to the extent that higher-
ranked organizational participants become more dependent on them. 30 It is
the situational characteristics which are of greatest interest to us, since the
effectiveness of the personal characteristics depend in large part on the
worker's organizational location.
Street-level bureaucrats command a degree of expertise, and indeed, of
deference, in some policy areas. (And they more or less display the personal
characteristics that enhance their influence.) But it is the discretionary role
of street-level bureaucrats and their position as de facto policy makers that
critically affect managers' dependence upon their subordinates. The sanc-
tioned discretion they exercise means that to demonstrate their own ability
and competence, managers are highly dependent upon subordinates without
being able to intervene extensively in the way work is performed.
Workers can punish supervisors who do not behave properly toward
them, either by refusing to perform work of certain kinds, by doing only
 Street-Level Bureaucrats as Policy Makers
minimal work, or by doing work rigidly so as to discredit supervisors. Police
officers, for example, can refuse to make vice arrests for a captain they
dislike, can refuse to take shortcuts through department regulations in order
to achieve results, or can rigidly and comprehensively enforce traffic or park-
ing regulations to the fury of an outraged public and the eventual embarrass-
ment of police officials. 31 Lower-level participants may also refuse to make
decisions that their superiors are formally obliged to make, at some costs to
superiors. Doctors who informally delegate dosage decisions to ward atten-
dants, 32 or judges who informally allocate sentencing decisions to probation
officers, are dependent on their hierarchical subordinates for the smooth
functioning of their jobs.
The relationship I have described between street-level bureaucrats and
managers has two primary characteristics. First, it is a relationship best con-
ceived in large part as intrinsically conflictual. The role of the street-level
bureaucrat is associated with client-processing goals and orientations
directed toward maximizing autonomy. Managers' roles in this context are
associated with worker-management goals directed toward aggregate
achievement of the work unit and orientations directed toward minimizing
autonomy. Second, it is a relationship of mutual dependence. Thus man-
agers typically attempt to honor workers' preferences if they are rewarded
by reciprocity in job performance. To a degree reciprocity will characterize
all working relations; in street-level bureaucracies, however, the resources
of lower-level workers are greater than those often possessed by subordi-
nates in other work contexts. Hence, the potential for reciprocity is greater.
This picture of workers and managers in street-level bureaucracies is sub-
stantially different from the one usually used to analyze problems of policy
making and implementation. Compliance with agency objectives may still be
the managerial problem, but it is complicated by the capacity of street-level
bureaucrats to resist organizational pressures with their own resources.
Some of these resources are common to public service workers generally
and some are inherent in their position as policy deliverers with broad
discretion.
When relationships between policy deliverers and managers are conflic-
tual and reciprocal, policy implementation analysis must question assump-
tions that influence flows with authority from higher to lower levels, and that
there is an intrinsic shared interest in achieving agency objectives. This situ-
ation requires analysis that starts from an understanding of the working con-
ditions and priorities of those who deliver policy and the limits on cir-
cumscribing those jobs by recombining conventional sanctions and
incentives.

25
 PART II

CONDITIONS OF
WORK

Introduction

STREET-LEVEL BUREAUCRACIES are consistently criticized for their

inability to provide responsive and appropriate service. The experience
of seeking service through people-processing bureaucracies is perceived by
enough people as dehumanizing that the phrase "human services" is often
understood as ironic by all but those who work under that label. 1
The persistence of rigid and unresponsive patterns of behavior results
from street-level bureaucrats' substantial discretion, exercised in a particular
work context. Like other policy makers, they operate in an environment that
conditions the way they perceive problems and frame solutions to them. The
work environment of street-level bureaucrats is structured by common con-
ditions that give rise to common patterns of practice and affect the direction
these patterns take. At base it is the shared situational context of street-level
work that permits generalization about these critical generic political and
social roles and the operational policies to which they give rise. 2
By definition, street-level bureaucrats work at jobs characterized by rela-
tively high degrees of discretion and regular interaction with citizens. Ordi-
narily, they also experience the following conditions in their work.

1. Resources are chronically inadequate relative to the tasks workers are asked to
perform.
z. The demand for services tends to increase to meet the supply.
3 Goal expectations for the agencies in which they work tend to be ambiguous,
vague, or conflicting.
CONDITIONS OF WORK

4 Performance oriented toward goal achievement tends to be difficult if not im-

possible to measure.
5 Clients are typically nonvoluntary; partly as a result, clients for the most part
do not serve as primary bureaucratic reference groups. 3

With the possible exception of the last item, to some degree these charac-
teristics follow from the definition of street-level bureaucrats. However,
they are worth elaborating because they call into play various behaviors in
which we are interested.
Note that these conditions of work may not always prevail. For example, it
may be that the welfare department in a small city with a stable, homoge-
neous, white. population is large enough to provide a relatively full range of
social services to recipients and perform with relatively clear objectives
derived from the relatively homogeneous political culture of the city. Em-
ployment and family patterns may be such that there is not a great deal of
movement on or off the welfare roles. In such a situation one might expect
social workers in that office to behave quite differently from the social work-
ers in a large, more heterogeneous central city.
The analysis presented here depends upon the presence of the aforemen-
tioned working conditions. If for some reason these characteristics are not
present, the analysis is less likely to be appropriate, although it is instructive
to understand why this is the case. If a legal services office encouraged its
staff to take only four or five cases at a time in order to maximize the quality
of preparation in each case, the lawyers would behave differently than if they
worked in an office with much higher demands.
Part II provides a detailed discussion of these work conditions and ana-
lyzes their contribution to the problem of providing discretionary social ser-
vices. Part III builds on this analysis to explore the mechanisms street-
level bureaucrats develop to cope with the difficulties and ambiguities inher-
ent in their work.
 CHAPTER 3

The Problem of Resources

Bureaucratic decision making takes place under conditions of limited time

and information. Decision makers typically are constrained by the costs of
obtaining information relative to their resources, by their capacity to absorb
information, and by the unavailability of information. 1 However, street-level
bureaucrats work with a relatively high degree of uncertainty because of the
complexity of the subject matter (people) and the frequency or rapidity with
which decisions have to be made. Not only is reliable information costly and
difficult to obtain but for street-level bureaucrats high case loads, episodic
encounters, and the constant press of decisions force them to act without
even being able to consider whether an investment in searching for more in-
formation would be profitable. In contrast, other bureaucrats at least some-
times may be able to calculate whether an additional effort should be made
to obtain a marginal increase in information. Indeed some organizations may
not be afflicted with resource scarcities at all. 2
Resource inadequacy is not only a theoretical consideration but a highly
practical one as well. This is because it appears to the public that resources
are manipulable and hence subject to calculated change. However, as we
shall see, this is not necessarily the case.
There are several ways in which street-level bureaucracies character-
istically provide fewer resources than necessary for workers to do their jobs
adequately. The two most important ways are the ratio of workers to clients
or cases, and time.
Street-level bureaucrats characteristically have very large case loads rela-
tive to their responsibilities. The actual numbers are less important than the
fact that they typically cannot fulfill their mandated responsibilities with
such case loads. Public defenders sometimes come to trial without having in-
terviewed their clients, or having interviewed them only briefly. Legal ser-

29
CONDITIONS OF WORK

vices lawyers, while responsible for perhaps 8o to 100 clients whose cases
they currently represent, may typically be working actively on only a dozen
or so cases. 3 Social workers are unable to make required home visits in
public welfare work and are so inundated with paperwork that they are
never without a backlog. 4
Lower-court judges are typically inundated with cases, often causing
delays of several months in providing defendants their day in court. Some-
how judges must accommodate pressures for speedier trials within the con-
straints offairness and equity. 5 In lower courts that hear misdemeanors the
volume is likely to be even greater than in felony courts, although the stan-
dards of justice are presumably the same.
For teachers, overcrowded classrooms (with meager supplies) mean that
they are unable to give the kind of personal attention good teaching
requires. High student-teacher ratios also mean that teachers must attend to
maintaining order and have less attention for learning activities.
For police officers, the obvious resource constraint is one of time-time to
collect information, time to act. A policeman breaking up a fight in a bar does
not have time to determine the initiating party and so must make a double
arrest. 6 Time to develop even a minimum degree of certainty appears to be
unavailable to officers who shoot innocent people because they feared the
consequences of what they say was the civilian's intention to reach for a
weapon. The problem of having to make a quick decision in life-threatening
circumstances is a major foundation of occupational conservatism in police
practice. Police reformers are unable to deal adequately with the heartfelt
claims of policemen that they require maximum discretion in order to pro-
tect themselves. This is the case even though their quick reactions may
evoke the very behavior they seek to defend themselves against, as in the
case of the citizen whose outrage results from the officer's apparent overreac-
tion to a situation.
Clearly, high case loads affect time for decision making. One observer of a
court that heard both felonies and misdemeanors observed that 72 percent of
the cases were handled in one minute or less. 7 Even when physical threat is
not particularly present, street-level bureaucrats must make quick decisions
because of the social reality that they are in the presence of clients who will
interpret indecision as incompetence or lack of authority, with consequences
for subsequent client interactions.
There are other organizational factors that affect the work of street-level
bureaucrats. An emphasis on housekeeping chores, such as filling out forms
or drawing up lesson plans, affects the amount of time available to clients. A
social worker who spends 6o percent of his or her time doing paperwork has

30
 The Problem of Resources
correspondingly less time for client interaction. 8 Support services (secre-
taries, clerks, investigators, receptionists) can affect the extent to which
street-level bureaucrats have time for clients. However, efforts to free
street-level bureaucrats of routine tasks so that they may attend to more im-
portant aspects of their work do not necessarily reduce the tensions as-
sociated with that work or improve the quality of interactions between work-
ers and clients. 9 The reasons for this are elaborated in the next section.
Street-level bureaucrats may also lack personal resources in conducting
their work. They may be undertrained or inexperienced. Police rookies, for
example, must undergo a long period of informal apprenticeship before they
are fully accepted and trusted by more experienced veterans. 10 The recent
law school graduate who enters the world of legal services is normally un-
trained both in interaction with clients and aspects of procedure that bear on
the kinds of problems clients are likely to face.
Street-level bureaucrats often experience their jobs in terms of inadequate
personal resources, even when part of that inadequacy is attributable to the
nature of the job rather than rooted in some personal failure. Some jobs just
cannot be done properly, given the ambiguity of goals and the technology of
particular social services. Determining whether or not a job is "proper" is
subject to uncertainties and implicit negotiation with others. (See chapter
10.) For example, there is no consensus on what techniques or approaches
substantially reduce rates of criminal recidivism, although social workers,
psychiatrists, and prison officials are supposed to be responsible for rehabilitat-
ing offenders. 11
From management's perspective street-level bureaucrats are resource
units to be applied to a task. But because of the nature of their tasks, workers
experience their work situations as individuals. One important respect in
which street-level bureaucrats respond as individuals is to the stress under
which they often work.
This is most obvious in the critical case of policemen, whose behavior
often can be explained only by their felt need to avoid danger. They con-
stantly work under the threat of violence that may come from any direction
at any time. Because the threat is unpredictable it exists constantly, although
the actual likelihood of threats materializing is quite low. 12 Police not only
guard against threat but they also characteristically conduct themselves so as
to determine whether threatening situations are likely to exist. They tend to
be lenient with offenders whose attitude and demeanor denote penitence
but harsh and punitive to those offenders who show signs of disrespect. 13
Indeed policemen often appear to test the extent to which an offender re-
spects police authority in order to determine whether he or she is likely to

31
CONDITIONS OF WORK

have an improper attitude and therefore be more likely to resist authority. 14

Corrections personnel confront similar situations and develop their own
mechanisms to establish control. 15
Psychological strain results from physical threat. In 1976, 1,500 of New
York City's 25,240 member police force were officially examined for psycho-
logical reasons, including alcoholism. 16 Law enforcement personnel are not
the only lower-level public employees for whom physical danger is a condi-
tion of work with which to reckon. According to a U.S. Senate subcommittee
report, between 1974 and 1975, 70,000 classroom teachers in the United
States reported serious injuries from physical assaults by students. This fig-
ure probably understates the situation considerably, since teachers are often
discouraged from reporting incidents of violence. Dr. Alfred Bloch, who
evaluated and treated more than 200 inner-city Los Angeles teachers who
had been assaulted while on the job, likens the psychological impact of these
experiences to battlefield trauma. 17
The threat of physical harm is the most dramatic aspect of the threat under
which street-level bureaucrats sometimes work. But virtually all interactions
between street-level bureaucrats and clients in some way focus on the work-
ers' needs to assert authority and gain client deference. This is most obvious
in the case of teachers who are instructed to first establish discipline as a
precondition to successful teaching. 18 However, it is generally characteristic
of street-level bureaucrats as well (see chapter g for further discussion).
Even without the threat of violence, street-level bureaucrats work in situ-
ations which tend to maximize the likelihood of debilitating job stress. One
recent study discovered significant correlations between relatively poor
mental health and three indicators characteristic of street-level work: re-
source inadequacy, overload (e.g. high case loads, overcrowded classrooms),
and role ambiguity.19
There is also a degree of stress in the limited extent to which street-level
bureaucrats feel themselves under scrutiny by authorities or others whose
negative evaluations might be harmful. This is illustrated well in the testi-
mony of George Kirkham, a law school professor who undertook to prepare
for and experience the police role. Kirkham draws attention here to the in-
teraction between feelings of pressure and the haste with which decisions
have to be made.

As a police officer, ... I found myself forced to make the most critical choices in a
time frame of seconds rather than days: to shoot or not to shoot, to arrest or not to ar-
rest, to give chase or to let g<r-always with a nagging certainty that others, those
with great amounts of time in which to analyze and think, stood ready to judge and
condemn me for whatever action I might take or fail to take. o
20
o 0

32
 The Problem of Resources
These points suggest that the salience of solutions to problems of resource
inadequacy varies not only with the demands on service and the resources
available, but also with the importance to an individual of deriving a satisfac-
tory solution to these problems.

Demand and Supply, or Why Resources Are Usually

Inadequate in Street-Level Bureaucracies

Most executives profess that their organizations do not have sufficient re-
sources or at least are hampered by resource constraints. Local governments
that tum back federal revenue-sharing funds are the exception to the general
rule that organizations perceive themselves as requiring and able to utilize
additional resources. Yet street-level bureaucracies, with certain other gov-
ernment agencies, chronically experience resource constraints. These agen-
cies are virtually never adequately provided for, and perhaps cannot be ade-
quately provided for. Why is this so?
A distinct characteristic of the work setting of street-level bureaucrats is
that the demand for services tends to increase to meet the supply. If addi-
tional services are made available, demand will increase to consume them. If
more resources are made available, pressures for additional services utilizing
those resources will be forthcoming.
The analogy to the development of traffic patterns on the Long Island
Expressway is compelling. In the name of relieving congestion during rush
hours on this infamous highway, traffic engineers added additional lanes.
But every additional lane, while marginally decreasing driving time to New
York City, induced more people to use the road. This additional traffic re-
stored the traffic jam that the new lanes had been designed to correct.
Utilization increased to meet the supply of road space until commuting
time reached the previous level. A new equilibrium was restored with the
same degree of congestion during rush hours, although with a higher volume
of traffic.
It has often been observed that utilization increases when public services
are expanded. Hospital emergency rooms become inundated because they
provide free medical care or easy access to care at a time when other health
resources, such as family physicians, are less and less available. 21 Mental
health centers discover that they may succeed only too well and evoke such a

33
CONDITIONS OF WORK

large response from citizens that they have to cut back service. 22 Additional
judges do not necessarily decrease court delays because with more judges
courts may be more tolerant of lawyers' delaying tactics, thus restoring the
lengthy time it takes to bring cases to trial. 23 Neighborhood multiservice-
center workers discover they have to abandon plans for systematic client
recruitment because the services of the new program are in such demand
that new cases prove overwhelming and case finding proves superfluous. 24
One dimension of service demand is quantitative. Public expectations of
and demands for certain public services increase over time. In the case of the
police, for example, society expects them to intervene in many more social
conflicts-interracial violence, black assaults on blacks, family disputes, ju-
venile justice-than was the case perhaps forty years ago. 25 Public expecta-
tions of personal health care services have become much greater over time,
spurred, to be sure, by developments in medical technology and health
insurance.
There is a great reservoir of demand for public services generally. In the
area of health, as in virtually every other area in which street-level bureau-
crats operate, there is no imaginable limit to the amount of health care the
population would seek and absorb if it were truly a "free good," available
without significant explicit or implicit costs. For some time the greater use of
physicians by middle-class people was thought by some to be simply a func-
tion of their greater willingness to use doctors, despite the fact that poor
health is inversely related to income. Studies have shown, however, that
when health care is made available at low cost there is little difference be-
tween people of different incomes in their search for medical attention. 26 It
is beyond the present capacity of health training institutions to train the
number of health professionals it would take to equalize the proportion of
health professionals throughout the nation.
More precise and narrowly focused examples of the latent demand for
street-level bureaucrats' services are available. At the end of the 196os for
example, it was reasonably estimated that to provide adequately for the legal
needs of the poor, 49,000 legal services lawyers and public defenders would
have to be employed. At the time there were perhaps 4,000 attorneys, both
volunteers and quasi-governmental employees, who were serving in those
capacities. 27
To comprehend the relationship between resources and practice, one
must understand the meaning of demand in public services. Demand is not
only part of a transaction between citizens and government but is also a
transactional concept. It not only requires a demander but also a more or less
encouraging supplier. It is meaningless to specify a level of demand unless

34
 The Problem of Resources
this is accompanied by a positive (if implicit) degree of receptivity. 28 In more
concrete terms, demand for services can be estimated but it cannot finally be
known in the abstract. It is a function not only of expressions of client prefer-
ences but also of government efforts to offer services and to record or ac-
knowledge client responses. Demand as an expression of desire for services
is a meaningful concept only if it is accompanied by explication of the extent
to which demand was sought out.
Overt expressions of demand for services tend to be more responsive to
changes in the perceived availability of services than in changes in the un-
derlying conditions that are commonly supposed to affect demand. In other
words, perceived availability of service "pulls" demand, not the other way
around. To illustrate: when cities introduce "gu" programs providing for
central reception of public safety telephone calls, an easily memorized tele-
phone number and publicity to instruct the public in its use, they discover
an untapped reserve of requests for public safety assistance. Indeed, New
York City at one point undertook a public campaign to urge people to use the
911 system less because the increased demand outstripped the city's ability
to respond. When Boston introduced its Little City Halls program, provid-
ing a municipal presence in the neighborhoods, requests for city services
through complaints increased by a third. 29
Service programs generally can increase the recorded demand on them-
selves by changing the nature of outreach, the location of offices, or public
information about the program. 30 Putative demand for services cannot even
take shape until the structures for receiving demands are in place. For ex-
ample, in Boston 4,000 students were assigned to classes for the retarded in
a recent year, whereas only 70 were classified as emotionally disturbed.
These figures might suggest that there was enormous demand for special ser-
vices for the retarded in comparison to the needs of emotionally disturbed
pupils. This situation would surprise most specialists in learning disabilities
because, to the extent that these labels have some objective referents, these
conditions tend to occur in about equal proportions in the population (about
2 percent). It would seem a reasonable inference that the disproportionate
demand for services for the retarded was not intrinsic among Boston youth
but rather was entirely a function of the availability of special classes for one
group and not the other and of the preferences of Boston school personnel
for assigning pupils to this category of disability. 31
The proposition that demand will increase to meet the supply applies
qualitatively as well as quantitatively. If there were a fixed clientele (we have
just argued there is not) clients would still demand more and improved ser-
vices, as the population has done historically. This might seem like the ideal

35
CONDITIONS OF WORK

toward which to strive, since conditions would be better for street-level

workers and clients. However, three possibilities need to be examined.
First, as mentioned earlier street-level bureaucracies usually must choose
additional services rather than improvement in services if they have slack
resources. Second, claims of qualitative improvements in the form of spend-
ing more time on each case are often spurious. Case loads are often infor-
mally divided into active and inactive categories. The inactive cases are often
not truly inactive but represent cases to which the street-level bureaucrat is
unable to attend in the ordinary course of the day. They are regarded as low
priority for reasons having little to do with the client but a lot to do with the
pressures on the workers. A social worker required to make more home
visits than he or she can possibly l'.rrange and still take care of more pressing
responsibilities, or a legal services lawyer with a large case load, only a por-
tion of which he or she can act on in the course of a week, have divided their
cases in such ways by necessity. 32 When additional workers enter these
agencies, they may reduce the formal case load by taking on a portion of each
worker's load. But presumably they will only be able to work with the same
number of clients as the other workers. Thus they will have the same active
case load, and everyone in the agency will have smaller inactive case loads.
More clients will be seen or served, but the amount of actual time spent with
the average client will not have improved. 33
Third, even when qualitative changes can be accomplished through case-
load reduction, the results are likely to be marginal. An additional teacher in
an elementary school of 300 pupils and an average class size of thirty would
theoretically reduce the average size of classes to twenty-seven. If this devel-
opment were reproduced system-wide it would mean a 10 percent increase
in teaching personnel. Yet its impact on the classroom work situation would
be welcome but marginal for teachers who still have to manage the learning
experiences of twenty-seven children.
With limited resources it might be desirable to add specialists rather than
relieve all classes equally. Yet the problem remains that the burden on gen-
eral classroom teachers would not be ameliorated. This is not to condemn
such developments, but only to raise the question whether even substantial
increases in public personnel budgets can reduce the work-load pressures
enough to make a difference in the way clients are processed if other condi-
tions of work remain the same.
When street-level bureaucracies do experience declining demand because
of population shifts and uneven age distributions, they encounter different
but equal difficulties in relieving case-load pressures. Consolidation or force
reductions tend to be administered so as to retain high individual case
 The Problem of Resources
loads. 34 Relieving case-load pressure may not directly translate into accept-
able bureaucratic behavior. 35 In particular, marginal reductions in case load
cannot be expected to result in visible improvements in practice. For ex-
ample, one would not expect teachers to differ substantially in the way they
handle disciplinary problems simply because their class sizes are reduced
from thirty to twenty-five. Pressures of inadequate resources may still be
responsible for questionable practice because: (1) variations in case loads do
not cross the threshold below which practice substantially improves; (z) case-
load pressures contribute to a milieu that remains even if conditions improve
slightly; and (3) the work context of street-level bureaucracies has several
components that interact with each other. Case-load pressure may interact
with other factors to determine behavior without necessarily varying directly
with the behavior in question. 36
A complication in providing service through street-level bureaucracies
comes about because the demand for service is sometimes unpredictable.
People who use or claim services cannot be counted upon to time their
needs to the exigencies of bureaucratic allocations. This is obviously the case
in emergency services, where the essence of adequacy is the capacity to deal
with the unexpected. An emergency room staff clearly cannot expect pa-
tients to appear at ten-minute intervals. Nor can they be expected to be pre-
dictable in their service demands, like patrons of a commuter railroad. It is
possible that an exceptionally affiuent street-level bureaucracy might be able
to handle unpredictable demands for service and provide superior service
during off-peak periods. But it is more likely that unpredictability combines
with pressing demand to impose considerable costs on the provision of ser-
vice. Workers may despair of ever catching up or otherwise getting out from
under the pressing burden of work. They may become insensitive to the
human dimensions of the job. 37 Certainly clients will bear many of the costs
of agencies not having the capacity to meet unpredictable demands. Long
and unexpected waiting times, broken appointments, short and hurried
treatment, are all costs that clients bear from the unpredictable (yet certain
to arise) system overload. 36
Street-level bureaucrats work in situations where the resource problem in
most cases is not resolvable. This is either because the number of people
treated by street-level bureaucracy is only a fraction of the number that
could be treated, or because their theoretical obligations call for higher
quality treatment than is possible to provide to individual clients, so that
slack resources are devoted to marginal improvements in quality. Conceiv-
ably, teachers could have classrooms of ten, legal services lawyers could
have case loads of fiA:een, and so forth. But the cost of providing service in

37
CONDITION S OF WORK

such ways are so high as to be politically out of the question except in the
case of tantalizing, carefully managed pilot or demonstration programs that
provide hints of what social welfare delivery might be like if service quality
were given primacy. Typically, if such programs succeed, their cost-be.1efit
ratios are regarded as unacceptable and pressures develop to provide service
in less costly ways.
This analysis of the demand-supply dilemma should not be taken as
counsel of despair. Public policy always requires consideration of the trade-
offs involved in providing additional resources for added benefits and incur-
ring additional costs. With added resources more people can be served, just
as more people can get to New York City from Long Island, although under
stressful conditions, by using an expanded Long Island Expressway. But ap-
preciation of the demand-supply dilemma in street-level bureaucracies does
suggest that the problem of the quality of service delivery is not likely to
yield easily to any imaginable resource increments. Other things being
equal, increased capacity results in reproducing the level of service quality
at a higher volume for any imaginable increase in resource availability. This
proposition is critical because it explains why the steady increase in re-
sources available to street-level bureaucracies in recent years has not re-
sulted in improvements in the perceived quality of client treatment. (Other
reasons include the fact that salary raises, which consume increases in
agency resource allocations, do not increase resources available to clients, al-
though they may help to maintain staff quality.) Further, it contradicts many
often self-serving perspectives on reform, which hold that additional person-
nel are the most important ingredients in responding adequately to citizen
complaints.
Thus street-level bureaucracies are often trapped in a cycle of mediocrity.
The better the program and the more responsive it is to the needs of the citi-
zens, the greater will be the demand for the service. This larger demand
forces the agency to limit service artificially or impose costs on clients in the
absence of a pricing mechanism that would otherwise ration service. The im-
po'ied cost of inferior quality or difficulty in receiving services continues
until, in the extreme, the agency is returned to the previous equilibrium of
indifference to client needs. The more successful the organization, the more
likely it is to encounter this dilemma.
It is generally characteristic of free government programs that demand
will increase to the point that goods or services can be provided. And in
response to this demand agencies providing the goods or services will im-
pose monetary or nonmonetary costs in order to limit demand effectively. 39
Street-level bureaucracies also develop rationing mechanisms to impose
 The Problem of Resources
costs for services but in doing so they face certain constraints. If indeed they
provide vital services-income to the indigent, public safety, education,
health care-then there are severe constraints on the sorts of rationing
mechanisms they can impose.
Street-level bureaucracies must not appear to be rationing services or
depriving social groups of their rights or entitlements. They must give evi-
dence of strenuous efforts to avoid reducing services. They may be asked to
"trim the fat," but never to reduce the quality of services or the quantity of
vital services. It is one of the best-kept secrets in government how agencies
can forever find fat to trim and nonessential services to eliminate, while
never affecting "vital programs" and "necessary services."
Agencies forced to reduce expenditures significantly, or otherwise limit
service provision, typically publicize the dire consequences of failure to spe-
cific groups or sectors to provide budgetary support for what they are doing.
(For example, the public university that announces it will be unable to admit
a freshman class.) But when it comes to making cuts they will typically at-
tempt to reduce services "across the board" rather than severely injure spe-
cific sectors of the population.
The demand-supply dynamics of street-level bureaucracies provide addi-
tional understanding of why they seem to be chronically understaffed.
The insatiable demand for services in basic social programs results in agen-
cies characteristically responding, and being forced to respond, by adding
services or clients, or extending existing resources under pressures of
budget stringency. The ethics of public service requires that they provide
more of what they do or give rather than utilize additional resources to
improve the resource-demand balance.
Sometimes agencies enjoy freedom from close scrutiny on spending. Then
they can add qualitative improvements or reduce case-load pressures, as edu-
cators have been able to do from time to time. Or, like the police, their ser-
vices may be in high demand, and they can ride public concerns to new
levels of spending for men and machines.
But for the most part street-level bureaucracies, as major components of
local public spending, cannot long escape close budgetary scrutiny and thus
become the targets of the taxpayers' revolt. In an inflationary era the budget-
ary convention of asking agencies to maintain effort on the basis oflast year's
allocation, while costs for goods, services, and salaries are rising, insures that
street-level bureaucracies will experience their resource position as chroni-
cally inadequate.

39
 CHAPTER 4

Goals and Performance Measures

Supervision and control provide guidance toward bureaucratic goals. Perfor-

mance measures offer feedback to adjust the system. The clearer the goals
and the better developed the performance measures, the more finely tuned
guidance can be. The less clear the goals and the less accurate the feedback,
the more will individuals in a bureaucracy be on their own. The ambiguity
and unclarity of goals and the unavailability of appropriate performance
measures in street-level bureaucracies is of fundamental importance not only
to workers' job experience, but also to managers' ability to exercise control
over policy.

Goals

Street-level bureaucrats characteristically work in jobs with conflicting and

ambiguous goals. Is the role of the police to maintain order or to enforce the
law? Is the role of public education to communicate social values, teach basic
skills, or meet the needs of employers for a trained work force? Are the goals
of public welfare to provide income support or decrease dependency? 1
Willis Hawley's comment on public education might well apply to virtually
every area of public service work: "[a]lmost every writer on educational ad-
ministration notes ... the problem of ... the multiplicity, ambiguity, and
diffuseness of goals. " 2
Public service goals also tend to have an idealized dimension that make
them difficult to achieve and confusing and complicated to approach. Goals
such as good health, equal justice, and public education, are indeed, as Mar-
tin Landau has observed, "more like receding horizons than fixed targets." 3
 Goals and Performance Measures
Agency goals may be ambiguous because the conflicts that existed when
programs were orginally developed were submerged. A typical mechanism
of legislative conflict resolution is to pass on intractable conflicts for resolu-
tion (or continued irresolution) at the administrative level. The much-
studied origins of the poverty program, for example, suggest that different
framers of this legislation had different ideas about the objectives of the Eco-
nomic Opportunity Act of 1g64, but could not resolve the inherent conflict at
the time the legislation was written. 4
Agency goals also may be ambiguous because they have accumulated by
accretion and have never been rationalized, and it remains functional for the
agency not to confront its goal conflicts. Goal conflict in welfare policy per-
sists not because analysts are unaware of ambiguity, but because there is
such fundamental disagreement among constituents of welfare policy that
Congress has never been willing to address and resolve the conflict directly. 5
Conflicting interests are deeply rooted and each side has marked out terri-
tory where it is influential. Each must be content with this arrangement
since neither side can prevail.
Another major source of ambiguity may be found in the uncertainty of
social service technologies. 6 When there are uncertainties over what will or
will not work, there is greater room for admitting and tolerating a variety of
approaches and objectives. In such a situation there is often a hunger for dis-
covering successful techniques and an apparent willingness to modify objec-
tives to suit the techniques. The speed with which new ideas in education
come and go primarily suggests a search for successful techniques, but also
indicates indecisiveness in objectives attributable as much to goal ambiguity
as to flexibility.
If goal conflict in street-level bureaucracies is fairly clear-cut (rather than
ambiguous), the conflicts characteristically have three sources.
1. Client-centered goals conflict with social engineering goals.
2. Client-centered goals conflict with organization-centered goals.
3 Goals conflict because street-level bureaucrats' role expectations are com-
municated generally through multiple conflicting reference groups.

CLIENT-CENTERED GOALS VERSUS SOCIAL ENGINEERING GOALS

A worker's concern for the client conflicts at times with the general social
role of the agency. This conflict comes to public attention perhaps most no-
ticeably in the case of police complaints that the public demands pursuit of
law and order in ways always consistent with norms of fairness and due pro-
cess. In the corrections area rehabilitation often conflicts with deterrence
and isolation of convicted criminals from the rest of society. In public welfare
CONDITIONS OF WORK

programs fostering the health and well-being of individual recipients con-

flicts with the goals of eliminating dependency and maintaining the attrac-
tion of low-wage work. Conflict between education oriented toward individ-
ual achievement and education oriented toward citizenship and discipline
also illustrates the tension between primarily focusing on individuals and
primarily focusing on social objectives.
Even in areas where one would think that client-centeredness would have
unambiguous primacy over other objectives, social engineering objectives
often dominate. The politics of legal services, for example, are currently
characterized by conflict over whether services should be provided on as
broad a basis as possible or whether they should be more concentrated to in-
sure that all poor people who secure legal services can count on having a law-
yer's services appropriate to their legal difficulties. In theory there is no
question that the professional relationship calls for the provision of services
as required. But this professional requirement is usually compromised in
organized practice for the poor. 7
At times client-centered goals primarily support social engineering func-
tions because of the symbolic importance of client-centeredness. Street-
level bureaucracies seek to gain client compliance either through the control
of resources that the client desires (utilitarian compliance) or, as in the case
of police and prisons, through force or the threat of force (coercive compli-
ance). 8 However, it is characteristic of a liberal society to show deference to
the norm of respect for the individual. Institutions are given license to orga-
nize and manipulate individuals only if they properly defer to this norm.
This is simultaneously a normative prescription for behavior and a dominant
element in social control.
Although little rehabilitation goes on in prisons the legitimacy of the
prison system depends in part on its claims to rehabilitative potential. Al-
though courts process people quickly, often with little time for complete
hearings on the merits of cases, rhetoric of guarantees for the "rights of the
accused" and the proposition that defendants are "innocent until proven
guilty" are powerful supports for a bureaucratized court system regardless
whether those injunctions protect defendants from injustice. Claims that ac-
tions are in "the best interest of the individual child" energize the rhetoric of
school personnel over an extremely wide range of educational policies, de-
spite considerable evidence that the way schools are structured precludes in-
dividual attention. Street-level bureaucracies that function through coercive
compliance systems such as the police still defer to the norm of respect for
individuals. Police legitimacy is strengthened by displays of regularity and
 Goals and Performance Measures
due process, although public interest in such symbols varies with the per-
ceived worthiness of the population accorded respect.
The distinction between coercive and utilitarian modes of compliance
often breaks down in street-level bureaucracies. Street-level bureaucracies
for the most part deal with nonvoluntary clients. This is obvious in the case
of the subjects of police arrests; it is less obvious in the case of social services
clients who appear to have the option of applying for benefits or refusing
them. However, if one takes the view that poor people have no alternatives
to seeking essential goods and services through public programs, then ana-
lytically the voluntary nature of their involvement in the system becomes
suspect. The alternative to volunteering to become a part of the welfare sys-
tem may be not eating. The alternative to going to a hospital emergency
room may be to neglect a serious health problem. If this is the case then to
call the client "voluntary" or the mode of client compliance "utilitarian"
becomes a bit scholastic.
Thus the norm of due process not only protects the rights of individuals
but it also legitimizes the effects the judicial and legal system has on people's
lives. Similarly, the norm offair treatment in public agencies combines with
the theoretical right to appeal to legitimize the actions of administrative
agencies. (Indeed, the development of standards for client treatment, rights
to appeal, and procedures for administrative regularity seem to develop in
proportion to client allegations of arbitrariness and unfair treatment. By de-
veloping procedural rules agencies may in fact protect the rights of some
clients, but they also gain legitimacy in continuing to act with most clients as
they did before.) Despite the relatively low status of the clients of many
street-level bureaucracies, the importance of deference to the norm of re-
spect for the individual is recognized by the considerable efforts made in
recent yectrs to invest clients of these systems with rights, although these
rights are often technical and remote rather than practical.
Some street-level bureaucracies such as schools and various rehabilitation
programs are both coercive and normative. Schools are coercive in that they
require attendance by law, they have strong sanctions against deviant behav-
ior, including expulsion, and they are the only institutions in the society
where parents can send children to learn basic skills. They are also norma-
tive in that they attempt to motivate children by socializing them to want to
participate in the system. A drug rehabilitation program is coercive when it
stands as an alternative to jail, but it also is normative because it cannot
succeed unless people are motivated to accept the orientation and discipline
of the program.

43
CONDITIONS OF WORK

In such settings the social engineering objectives may simply be to place

people in the hands of skilled professionals. There will be inevitable conflicts
between the client focus of professional norms and the dominant framework
of the institutions in which the professionals work. A teacher may, as a pro-
fessional, be oriented to individual evaluation and assistance, but the school
as an organization may have as competing orientations discipline, citizenship
training, and socialization to norms of general behavior. 9
In addition, the operating necessities of the organization will likely inter-
fere with providing professional services. The example of the court psychia-
trist illustrates this familiar observation. Psychiatrists are trained to render
highly personalized and individualized services through interaction with
clients in a sustained period of consultation and intervention. Yet the court
psychiatrist is often placed in the absurd position of having an extremely brief
period to diagnose and label people referred by judges anxious to avoid con-
signing defendants to the harsh and counter-productive penalties of the
prison system. Court psychiatrists are chronically misused because the social
function of psychiatric referral by judges outweighs protection of the psychi-
atric milieu.

CLIENT-CENTERED GOALS VERSUS ORGANIZATIONAL GOALS

Street-level bureaucracies encounter conflict and ambiguity in the tensions
between client-centered goals and organizational goals. This distinction
echoes some of the considerations presented above and in some ways is con-
tinuous with it. The ability of street-level bureaucrats to treat people as indi-
viduals is significantly compromised by the needs of the organization to pro-
cess work quickly using the resources-at its disposal.
The fundamental service dilemma of street-level bureaucracies is how to
provide individual responses or treatment on a mass basis. Martin Levin
calls attention to this conflict in his study of criminal courts. "At the most
general level the motives of the judges are primarily two: the defense and
maintenance of their traditionally high degree of discretion in criminal court
. . . " permitting them to tailor decisions to the needs, "and successfully
coping with their court's case loads." He asserts: "the judges tend to empha-
size processing their case loads as an end in itself 'rather than a means for
achieving other goals." 1o
This is a classic example of goal displacement. Effective management of
the process preempts the purposes for which the process was developed.
The study of street-level bureaucrats may be seen as a study in goal displace-
ment when the norm of individual client orientation becomes subordinate to
the needs for mass processing. The typical conflicts here are individual client

44
 Goals and Performance Measures
treatment versus routinization and mass processing, and response to the
needs of individual clients versus efficient agency performances.
These dilemmas are related to the public nature of the programs. Just as
agencies distributing free goods must develop mechanisms to ration their
allocation, political systems must place limits on the demands that organiza-
tions can make for additional resources. If, indeed, demand will increase to
equal the supply then the inherent impulses organizations display toward
growth 11 will lead to increasing organizational scope even though, as sug-
gested above, the quality of service cannot be expected to improve with
growth.
Suggestive evidence for this proposition is encountered in the alarm that
public officials express when they discover that benefit programs are open-
ended. The food stamp program, for example, caused a great deal of official
consternation because it entitled many more people to benefits than was
originally anticipated, resulting in unexpected costs difficult to control ex-
cept through categorical cutbacks. 12 Constant harping on the error rate in
public welfare or allegations of doctor abuses of medicare claims serve to
remind the public and the agencies in charge of these programs that uncon-
trolled growth in government spending is not officially acceptable. Thus, ex-
cept in rare instances, such as a new service program in search of a clien-
tele, 13 street-level bureaucracies are under continuous pressure to realize
the public objectives of efficiency and cost effectiveness. Pressures will be
more or less explicitly articulated depending upon the political climate and a
variety of other factors.
Like other organizations, street-level bureaucracies are always con-
strained by resource limitations. However, it is important to understand that
this constraint is experienced as a source of tension within street-level bu-
reaucracies at the operational level. Street-level bureaucrats must find a way
to resolve the incompatible orientations toward client-centered practice on
the one hand and expedient and efficient practice on the other.

GOAL CONFLICTS AND ROLE EXPECTATIONS

Finally, goal conflicts and ambiguity arise from the contradictory expecta-
tions that shape the street-level bureaucracy role. Generally role theorists
locate the origin of role expectations in three sources: in peers and others
who occupy complementary role positions; in reference groups, in terms of
which expectations are defined although they are not literally present; and in
public expectations where consensus about role expectations can sometimes
be found. 14 To the extent that these. sources of role expectations differ signif-
icantly one would expect street-level bureaucrats to encounter role conflict

45
CONDITIONS OF WORK

and ambiguity (note that goals are one dimension of role construct). There
are at least three ways in which the complicated structure of street level bu-
reaucrats' role expectations contributes to goal ambiguity and conflict.
First, to the extent that public expectations affect street-level bureaucrats
there is often considerable disagreement about what street-level bureau-
cracies should primarily do. Street-level bureaucrats within limits may de-
fine the ways in which they will pursue their objectives. But often commu-
nity opinion is diffusely apprehended, creating role conflict. In different
cities and within the same city some communities may expect police to en-
force the law vigorously and impartially and to have a strong legalistic orien-
tation, while others are content to have the police primarily focus on main-
taining order. Some parents want schools oriented toward achievement in
basic skills, while others want schools oriented toward developing commu-
nity values and vocational training. The punitiveness of welfare administra-
tion in some cities and the liberal orientation of other welfare offices demon-
strate clearly how the same statutes can spawn agencies with vastly different
bureaucratic cultures. 15
These differing perspectives on the purposes of service contribute to goal
uncertainty and lead to the following hypotheses. To the extent that com-
munities are indifferent to the nature of bureaucratic policy or fail to express
their views in politically salient ways, street-level bureaucracies will perform
with internally generated objectives. Conversely, the stronger community
sentiment is concerning proper bureaucratic behavior, the more street-level
bureaucracies will respond to community orientations. The more heteroge-
neous community sentiments are, however, the more street-level bureau-
cracies will experience goal conflict.
Major urban conflicts have resulted in recent times over such diverse
perspectives on public services. Should schools promote appreciation of eth-
nic and racial heritage or should they be devoted to the traditional role of
homogenizing and Americanizing urban newcomers? Should public housing
managers attempt to integrate their projects racially or should they try to
maximize the security and comfort of long-term project residents? Is dis-
cipline and punishment more effective for inducing studious attitudes among
pupils or is a liberal and flexible approach more appropriate? These are ques-
tions that would be answered quite differently by different subpopulations of
the country's largest cities. 16
Racial conflict brings issues of divided community sentiment into sharp
focus. Black community residents differ from whites in their law enforce-
ment priorities, the desirability of police intervention in certain areas, ap-
propriate role models for their children in schools, and so on. Moreover, res-
 Goals and Performance Measures
idents of black communities themselves differ in priorities; for example,
whether they would prefer high-intensity police protection in order to
reduce crime or whether they regard saturation patrolling as community
harrassment .
A second dimension of role conflict or ambiguity stems from the significant
role of peer groups in establishing role expectations. For street-level bureau-
crats, peers are fellow workers (although generally peers can be otherwise,
e.g., social peers, family peers, etc.). Only work peers fully appreciate the
pressures of work and the extent to which street-level bureaucrats experi-
ence the need to have goal orientations that are consistent with resolving
work pressures. The greater the strain between various goal expectations,
and the smaller the zone of indifference in which street-level bureaucrats
operate, the more peer support is critical for sustaining workers' morale. 17
The subject of peer support is discussed extensively in analyses of the
police, the street-level bureaucracy that is perhaps the most controversial
and the most subject to conflicting goal expectations. Police must perform
their duties somewhere within the demand for strict law enforcement , the
necessity for discretion in enforcemen t actions, and various community in-
terpretations of proper police practice. They must accommodate the con-
straints of constitutional protection and demands for efficiency in mainte-
nance of order and crime control. They must enforce laws they did not make
in communitie s where demands for law enforcemen t vary with the laws and
with the various strata of the population. 18 Police may perceive the public in
these communitie s as hostile yet dependent. Police role behavior may con-
flict significantly with individual value preferences and the behavior and
outlook of others with whom police must work, particularly judges. Police
are expected to be scrupulously objective, impartial, and upstanding, pro-
tective of all segments of society even while society does not protect all seg-
ments of itself. 1e
A third dimension of the construction of street-level bureaucrats' role ex-
pectations concerns the role of clients. Clients are not a primary reference
group of street-level bureaucrats. They do not count among the groups that
primarily define street-level bureaucrats' roles. 20
This is not to say that children are unimportant to teachers or that litigants
and defendants are unimportant to judges. But these people do not primarily
or even secondarily determine bureaucratic role expectations. Work-relate d
peer groups, work-related or professionally related standards, and public ex-
pectations generally are much more significant in determining role behavior.
Recognizing their weak influence in defining workers' roles, some client or-
ganizations have demanded inclusion in the constellation of bureaucratic ref-

47
CONDITIONS OF WORK

erence groups. 21 Street-level bureaucrats' resistance to client demands may

be understood by recognizing that clients are not part of their reference
group constellation. The fact that many street-level bureaucrats provide
client services or are required to interact with clients in a helpful manner in
no way implies that they think that clients should have a say in the nature of
street-level practice. Indeed, the organizations that collectively articulate
the perspective of street-level bureaucrats, such as teachers' and patrolmen's
associations, have fought vigorously to keep their arenas free from citizen
involvement.
Yet the matter is hardly simple, as community disputes over greater client
participation indicate. Since street-level bureaucrats generally pay lip ser-
vice to clients' opinions, and as individuals they no doubt often do respect
client views, there are internal and professional pressures that incline them
toward accepting clients' judgments and soliciting their respect and con-
sent. The cauldron into which clients' demands for a voice in policy have
been placed has brewed some of the most vigorous community conflicts of
the last fifteen years.
Overt conflict is not the only result of role ambiguity. Lack of clarity in
role expectations has been found to impair personal action and reduce work-
ers' effectiveness. 22 Thus role ambiguity affects individual performance as
well as organizational direction.

Performance hleasures

Job performance in street-level bureaucracies is extremely difficult to mea-

sure. The many implications of this statement include the facts that these
agencies are not self-corrective, and the definition of adequate performance
is highly politicized.
For some purposes bureaucracy itself may be defined in part as a large or-
ganization whose output cannot be evaluated through market transactions.
Such a definition distinguishes bureaucracies from business organizations,
whose behavior is in a sense assessed through profitability. 23 To say that out-
put cannot be evaluated through market transactions draws attention to the
fact that bureaucracies characteristically cannot have their work evaluated
through an organic social process such as that which is symbolized by the
summary on a balance sheet. While in theory a market-oriented organization
can learn when it is succeeding or failing through the inexorable realities of
 Goals and Performance Measures
profit and loss, bureaucracies receive no similar messages. Hence the mea-
surement and evaluation of performance-the governance of perfor-
mance-is critical.
Difficulty in evaluating performance may be characteristic of bureau-
cracies generally but it is particularly endemic in street-level bureaucrats'
work. When the output consists of services provided or the validity of discre-
tionary decisions made, it is extremely difficult to oversee or scrutinize these
decisions if standards of quality are at issue. The following observations con-
tribute to understanding the problem of performance measurement. 24
Goal ambiguity, intrinsic to street-level bureaucracies, affects perfor-
mance measurement. Such goals as developing a well-educated citizenry or
securing public safety are ephemeral. But putting these general goals into
operation is a practical matter, since we evaluate achievement in terms of
goals. As we have seen, there is no agreement in society about objectives of
public education and public safety forces. How then to operationalize ambig-
uous objectives? The extraordinary attention practitioners and theorists in
public administration recently have placed on clarifying organizational ob-
jectives testifies as well as anything to the general importance of goal clarity
for organizational evaluation.
Still another reason that street-level bureaucrats' performance often
eludes effective evaluation is that there are too many variables to take into
account to make evaluation realistic. It is not only that human beings are
complex and that a metric of correct responses is inappropriate. Equally im-
portant, there is rarely any way to determine on a regular basis what would
have happened to clients in the absence of intervention. A job training pro-
gram may have a 40 percent placement rate, but to assess its effectiveness
one would want to know the employment potential in the trainee group, and
what the employment rate of participants would have been without the pro-
gram. If the program drew its participants from the most work-resistant part
of the population, then a 40 percent rate might be extremely good. How-
ever, if participants were drawn from a work-oriented part of the popula-
tion-that is, if the population were "creamed"-then its placement rate
might be judged inadequate. Placement rates can only be evaluated if one
knows the quality of the material with which programs started.
For example, in the Upward Bound program in the 1g6os evaluation of in-
dividual projects ultimately proved impossible because there was no way to
determine whether programs that got a high proportion of their students
into college started with high-risk or low-risk students. The program direc-
tors were encouraged to admit high-risk students but no one could tell from
testing or background information available on students who was or was not a

49
CONDITIONS OF WORK

high risk. Being poor, black, and a student at an inferior high school was
regarded as high risk enough. Programs often admitted students with rea-
sonable grades, neglecting students who might have been even higher risks.
If those programs succeeded did they succeed because they selected stu-
dents prudently, or because they ran effective programs? Moreover, evalua-
tion would have to consider changes in the social or economic environment.
Changes in national employment trends or civil rights orientations might af-
fect program success in ways not attributable to the program.
To some degree public deference to street-level bureaucrats' autonomy in
decision making is also characteristic. This deference is a defining aspect of
professionalism and has some applicability in all the areas in which street-
level bureaucrats work. While public deference is related to the needs for
discretionary judgment, goal ambiguity (leaving goals to be defined by pro-
fessionals), and decision complexity, it has a meaning of its own, particularly
as public bureaucracies attempt to develop performance measures as a pre-
condition to increasing bureaucratic control.
Street-level bureaucrats tend to perform in jobs that are freer from super-
visory scrutiny than most organizational jobs, and work norms prevailing in
these jobs minimize such scrutiny. This may be because of supervisors'
willingness to respect professional claims (for example, teachers in most
schools are rarely visited in classrooms by principals, and then only with
enough notice so that performances can be staged). Or it may be because
workers expected to exercise discretion require some freedom from super-
visory control (for example, policemen whose supervisors leave them alone
as long as they seem to be performing adequately). 25 Whatever the sources
of this freedom, it contributes in itself to the problems of measuring perfor-
mance, particularly since peer evaluation is one of the ways to achieve ac-
countability in work quality.
Notwithstanding these difficulties, however, bureaucracies do establish
standards and measure workers' performances against these standards. For
example, policemen typically are asked to make a certain number of arrests
per month. Social workers are asked to maintain a certain monthly intake
and case-closing rate. But these measures are only problematically related to
public safety, or to clients' ability to cope with problems that are in part the
objectives of these interactions. And they have nothing to do with the appro-
priateness of workers' actions, or the fairness with which they were made,
the net results of which determine the rates on which workers are judged.
Not only are such standards problematically related to goals, but it is not
even apparent whether measured increases or decreases signal better or
worse performance. The best illustration of this consideration is the problem

so
 Goals and Performance Measures
of inference from crime statistics. Do increases in arrest rates signal im-
proved police performance? Or do they signal deteriorating police perfor-
mance, indicating an increase in criminal activity and thus in the number of
criminals available to catch? Actually, changes in arrest rates may indicate
neither, reflecting rather changes in the focus of police patrol. Do decreases
in the welfare rolls signal improved bureaucratic performance? They do only
if improvement is defined as having fewer people on welfare, the economy
that creates the need for welfare remains stable, and welfare agencies do
not change their practices affecting clients' willingness to apply. These are
but specific illustrations of the observation that agency-generated statistics
are likely to tell us little about the phenomena they purport to reflect, but
a great deal about the agency behavior that produced the statistics. 26
Despite the difficulties of performance measurement, street-level bureau-
cracies do seize on some aspects of performance to measure. They tend to
seek reports on what can be measured as a means of exercising control. In
tum, the behavior of workers comes to reflect the incentives and sanctions
implicit in those measurements.
The relationship between performance measures and behavior was per-
haps first highlighted by sociologist Peter Blau when he observed that when
the employment agency he WllS studying began to be evaluated in terms of
its placement rate, employment counselors shifted the focus of their work to
the more easily employed at the expense of those more difficult to place. 27
This illustrates the general rule that behavior in organizations tends to drift
toward compatibility with the ways the organization is evaluated.
Street-level bureaucracies also measure the training and experience of
employees as a way of assessing quality. These surrogates represent qualities
that are hypothetically associated with good performance. Thus a teaching
faculty is assessed in terms of its educational background although the gradu-
ate training in question is only problematically related to effective teaching.
Street-level bureaucracies often depend on the experience or training of
their workers as signs of quality service, although it is not clear that more
training or experience is associated with doing a better job. Low-income
parents often express concern that experienced teachers exercise seniority
rights and migrate to schools in -affiuent neighborhoods. But experience is
problematically related to effective teaching. The less-experienced teacher
may be more interested, energetic, ambitious, more recently educated and
familiar with new techniques, and more sympathetic toward a low-income
clientele. (Of course, lacking good performance measures we rarely have an
opportunity to test the relationship between the surrogate measures and
their implications for performance.)

51
CONDITIONS OF WORK

The organization itself develops surrogate measures of performance on

which it evaluates and is evaluated. Case loads, arrest rates, home visits, and
worker/client ratios are all measures of work activity that serve public agen-
cies in the absence of performance measures. These surrogate measures
then become reified and guide future performance. 28
Organizations tend to measure what they can readily quantify without in-
truding on workers' interaction with clients. Organizational attention focuses
on two major considerations. First, a great deal of attention is paid to the way
the worker spends his or her time. Classroom control and other demon-
strations of discipline are important criteria in teacher evaluations largely
because these are aspects of the teacher's classroom that can be observed
without intrusion. 29
Since the determining factor in accomplishment of the goals of the school is the con-
tact between students and teachers in the institutional framework of the school, the
main control that the organization has over the action of teachers in the accomplish-
ment of the school's goals is the enforcement of rules concerning teacher behavior.
Thus, the administrator can measure teacher compliance with these norms rather
than teacher effectiveness since, unfortunately, for many administrators the two are
synonymous. 30

Organizations may develop surrogate indicators of performance and qual-

ity, but workers accommodate themselves to these measures and remain in-
dependent of organizational control. Since the quality of their accommo-
dations remains unscrutinized, police can meet traffic ticket quotas in a single
day or reduce serious crime by recording burglaries as larcenies. 31 Housing
inspectors can appear to increase their productivity by inspecting more pre-
mises, but this at the expense of lowering their standards and reducing their
time spent per inspection. 32 Street-level bureaucracies attempt to promote
the validity of surrogate measures to the general public in an effort to appear
accountable through performance standards. Although they currently make
great efforts to develop information systems to give the impression that they
actively seek to increase productivity, there are really few valid statistics
where the quality of performance is at issue.
Albert 0. Hirschman has usefully pointed out that thef'e are two general
ways in which organizations can be self-correcting. Members or consumers
may withdraw their organizational support and by so doing signal to man-
agers that something is wrong. Alternatively, they can express their con-
cerns or objections and by so doing help organizations to correct them-
selves. 33 However, citizens' ability to affect organizations by withdrawing or
speaking up depends upon the clarity with which they receive information
about the organizations. 34

52
 Goals and Performance Measures
An important problem of public bureaucracies generally and street-level
bureaucracies particularly is that clients do not receive the kind of informa-
tion that would permit them to compare or assess their treatment. Nor can
they compare the treatment they receive this year with the treatment ex-
tended to clients in other years, or compare the performance of their agency
with similar agencies elsewhere. Citizens in general and poor people in par-
ticular will resign themselves to inferior levels of service if they have nothing
with which to compare their experiences and have no basis for thinking that
they deserve any better. Their frame of reference, if any, is experiential. But
the isolation of most clients from each other makes it difficult to interpret ex-
periences effectively and makes clients highly subject to street-level bureau-
crats' definition of their situation.
Recent experiences with educational voucher experiments illuminate the
importance of the absence of performance measures in street-level bureau-
cracies. First, teachers were very resistant to comparison to other teachers
and to the staffs of other schools. Second, parents rarely exercised their op-
tion to shop around for the best school, tending instead to support the con-
ventional schools. From the standpoint of the present analysis one could
hardly expect parents to exercise options when they had no basis for under-
standing which schools performed better. Whatever problems they may
have had with their schools parents may have attributed to difficulties within
themselves or their children. In the absence of solid information on alterna-
tives the security of the known quantity was in any case generally chosen
over the unknown quantity. Significantly, in the experiments in which
school systems were able to differentiate the programs of each school,
parents did exercise considerable options. 35
In summary, the inability to measure street-level bureaucrats' perfor-
mance has widespread implications for controlling the agencies. Supervisors
and agency directors can discipline workers, but not to the point of closely
guiding workers' activities toward agency preferences unless they can moni-
tor performance and determine who is or is not measuring up. Nonetheless,
surrogate performance measures are developed to provide the agencies and
the public with control tools, even if the tools are not quite appropriate and
may even be counterproductive to the purposes. These measures do affect
workers' behavior, although not necessarily in the direction favored by the
agency or the public. And street-level bureaucrats, in recognition of the im-
portance performance measures have to limiting their autonomy, actively
resist their development and application.

53
 CHAPTER 5

Relations with Clients

Nonvoluntary Clients

Clients in street-level bureaucracies are nonvoluntary. This point is obvious

in coercive public agencies such as police departments, but it also applies
when the coercive dimensions of the relationship between the agency and
the client are less clear. This is because street-level bureaucracies often
supply essential services which citizens cannot obtain elsewhere. Govern-
ment agencies may have a monopoly on the service, clients may not be able
to afford private services, or they may not have ready access to them. Poten-
tial welfare recipients in a sense "volunteer" to apply for welfare, for ex-
ample, but their participation in the welfare system is hardly voluntary if
they have no income alternatives.
Where government does not monopolize an essential service it often pro-
vides the only such service available to the poor. Health care and legal ser-
vices, for example, can be obtained privately but only at relatively high cost.
The cost of obtaining private assistance in these areas is so great, relative to
income, that poor people are forced to seek assistance through public agen-
cies or not to seek assistance at all. The poorer the person, the more he or
she is likely to be the non voluntary client of not one but several street-level
bureaucracies. Relatively affiuent people can seek out a private doctor in-
stead of a public clinic. Middle-class families dissatisfied with their school
systems can change residences. These options are less available to the poor.
This relationship between poor people and public agencies provides grounds
for concluding that poor people receive a qualitatively different kind of treat-
ment from the state. Indeed, we might predict that public agencies with

54
 Relations with Clients
poor clients provide different treatment in comparison to those serving more
affluent people entirely on the basis that poor people are much more depen-
dent on these agencies. 1
What is the meaning of saying that the clients of street-level bureaucracies
are nonvoluntary? What difference does it make and why do we elevate this
consideration to the status of a primary working condition?
If street-level bureaucracies have nonvoluntary clients then they cannot
be disciplined by those clients. Street-level bureaucracies usually have noth-
ing to lose by failing to satisfy clients. They will try to manage a large volume
of complaints and undoubtedly seek to minimize the extent to which they
are perceived as difficult to deal with or unresponsive. But managing com-
plaints successfully is a far cry from changing policy in response to consumer
dissatisfaction. Yet, as indicated in the previous chapter, receiving com-
plaints and correcting policy in response to them is one of the few ways orga-
nizations can learn from clients. 2
Sometimes street-level bureaucracies are even rewarded for reducing
their clientele. Public welfare agencies in the United States come under crit-
icism for their large number of clients, although a public agency that
reached less than half the eligible population in other contexts might be
regarded as a colossal failure. At other times street-level bureaucracies are
indifferent to the loss of clients or client dissatisfaction. Partly this is due to a
proposition developed earlier. If demand for services is practically inex-
haustible relative to supply, then the fact that some clients are disaffected by
the quality or level of service means only that their places are taken by
others who need the service and are willing to accept the costs of seeking it.
Even in situations in which one would assume that street-level bureau-
cracies would suffer from client losses, the bureaucracies often seem rela-
tively indifferent. Consider the case of public schools in central cities that
have lost white, middle-class students and failed to retain lower-class stu-
dents. Although school revenues are based on average, per-pupil attendance
these population losses usually have little effect on school practices.
Perhaps there are ways to explain this lack of responsiveness. The tenure
of the more experienced teachers, for example, may mean that they do not
consider their jobs threatened by declining school enrollment. Or perhaps
school officials think that they are not capable of combatting the attractions of
suburban and parochial schools or of influencing achievement levels among
drop-outs. Whatever the full explanation of schools' unresponsiveness to
declining enrollments, 3 they place surprisingly little emphasis on programs
to retain pupils and on public relations designed to promote their strengths

55
CONDITIONS OF WORK

and enhance their long-run budgetary status. Instead schools fight for their
budgets within the arena of local-interest group politics without great em-
phasis on consumer satisfaction.
That clients are nonvoluntary has significant implications not only for the
direction of public services as a whole but also for the quality of interactions
between street-level bureaucrats and clients. Primarily, this is because non-
voluntary clients cannot avoid or withdraw from encounters with workers.
Where both parties are free to continue the interaction or leave it, partici-
pants will set limits to the costs they will accept before ending the rela-
tionship. If the encounter is instrumental, that is, if each participant wants
something from the other, they will continue to pursue their objectives
within the relationship so long as they value the objectives more than the
cost of seeking them. This permits a wide range of implicit bargaining tac-
tics, particularly if both parties have a stake in maintaining the relationship. 4
However, if one of the parties does not enter the relationship voluntarily
or must sustain the relationship because a highly desired good for which
there is no alternative is controlled by the other person in the encounter, the
nature of the interaction changes. The costs that the non voluntary person in
the interaction will sustain become much higher. Indeed, the less voluntary
the interaction, the less useful it is even to understand the interaction in
terms of limits to the costs people will accept, because clients cannot easily
withdraw.
Street-level bureaucrats can impose costs of personal abuse, neglectful
treatment, or inconvenience without necessarily paying the normal penalty
of having the other party retaliate. When medical personnel refer to patients
as "garbage," "scum," "liars," "deadbeats," and so forth, there is a tempta-
tion to say that this is a reaction to the moral superiority they feel over lower-
class people. 11 However, neglect and abuse of patients is a function of the
nonvoluntary nature of the association of clients with patients, and not
strictly of bureaucracy or class discrepancies. This is suggested by the obser-
vation that doctors in private practice can also neglect patients if there is a
restricted supply of professionals, but they become much more solicitous
when patients have medical alternatives on which to draw. 6
In the elliptical euphemism used by several street-level bureaucracies,
they need not worry that clients will"elope." If clients refuse to continue in-
teracting with street-level bureaucracies, the fault may always be attributed
to the client. "Escapees," "dropouts," "incorrigibles," and "socially disorga-
nized" are labels that imply that the exit of the client is attributable to a
defect of the client.
 Relations with Clients

Conflict, Reciprocity, and Control

The nonvoluntary nature of clients helps explain why they are not among
street-level bureaucrats' primary reference groups (see chapter 4). But this
does not mean that clients are helpless in the relationship. Street-level bu-
reaucrats in a sense are also dependent upon clients. Clients have a stock of
resources and thus can impose a variety of low-level costs. This is because
street-level bureaucrats must obtain client compliance with their decisions,
particularly when they are evaluated in terms of clients' behavior or
performance.
Order in a prison is a function of adjustments made by guards in exchange
for prisoners' general compliance with regulation. 7 So it is with most social
organization. To do their work smoothly police officers must obtain the con-
sent of suspects they apprehend. Teachers must secure pupils' cooperation
before they can begin to teach. Social workers must obtain the compliance of
welfare recipients in case determinations or confront time-consuming ap-
peals. The child who refrains from asking a question after a certain point out
of fear of making the teacher angry, the traffic law violator who fails to con-
test what he regards as an officer's unfair decision, the law client who, al-
though vaguely troubled, is too confused to raise questions about her case,
are all actively cooperating in the interactions.
For the most part, except in the more coercive bureaucracies, clients give
their consent because (sometimes in combination) they accept the legitimacy
of the street-level bureaucrats' position and decision, anticipate that dissent
would not be productive, or consider themselves favored by the decision or
action taken. Most encounters with bureaucracy appear to be characterized
by the consent of clients, but the structure of choices available to clients
limits the range of alternative behaviors that they consider realistically avail-
able. In short, clients' consent is continuously being managed by public
agencies.
Street-level bureaucrats are not required to command. Clients control
themselves in response to the superior power of the workers. This is not to
suggest that clients are docile because swift retaliation would result from
noncompliant behavior. Rather, compliance in most street-level bureau-
cracies may be said to result from the superior position of the workers, their
control over desired benefits, and their potential capacity to deny benefits or
make their pursuit more costly.
Compliance also results from the milieu, which comprehensively cues
clients concerning behavioral expectations. Readers lower their voices in

57
CONDITIONS OF WORK

libraries, defendants talk respectfully to judges, patients wait quietly for doc-
tors, children obey teachers, not directly because of the imminence of retali-
ation but because they have a diffuse appreciation of "proper" modes of be-
havior and a diffuse awareness that deviance from these norms may be
punished. 8
Nonetheless, street-level bureaucrats sometimes do display behavior that
strongly suggests this inference is warranted. Street-level bureaucrats in-
deed reprimand or otherwise sanction deviance from acceptable standards of
client behavior. They dominate their interactions with clients. They cue and
otherwise teach clients to behave "properly." They structure work patterns
to maximize control over clients independent of any policy objectives.
The need to control clients is a requirement evident in many work areas,
not just street-level bureaucracies. Relationships are always reciprocal to
some degree. If one party seeks to control the other, the second party may
increase the costs of the first party gaining or exercising control, even if the
first is unquestionably more powerful. This observation, which has universal
applicability from guerrilla warfare to concentration camps, takes particular
shape in street-level bureaucracies in several respects.
First, street-level bureaucrats characteristically are pressed with heavy
case loads and demands for quick decisions, so that clients can impose salient
costs merely by taking workers' time. Since time may be fairly cheap for
clients, or their needs high relative to the value they place on their time,
clients potentially have a store of resources with which to affect their rela-
tionships with street-level bureaucrats.
Second, street-level bureaucrats are characteristically constrained in the
resources they can employ in obtaining client compliance. These constraints
consist of professional and bureaucratic standards of fairness and due process
that to some degree place limits on what can or cannot be done to or with
clients (notwithstanding the most outrageous tales of exceptions to the con-
trary). They are also constrained by social norms of proper behavior toward
other people and by recognition that power should be accompanied by re-
sponsibility, particularly when clients are identifiably (indeed defined as)
socially or economically needy. This point is emphasized not because street-
level bureaucrats are absolutely constrained from abusing their positions,
but because what needs to be explained is the mobilization of control in com-
bination with constraints against excessive manifestations of power. Modern
bureaucracies which are too heavy-handed lose their legitimacy if their of-
fenses are publicized. Moreover, they are ultimately inefficient if they
require significant force to assure adequate client control.
Third, there is an extent to which clients' satisfaction or performance is

ss
 Relations with Clients
important to street-level bureaucrats. Successful intervention, expressions
of gratitude, and changes in behavior in the desired direction are valued by
street-level workers whether or not these developments are reasonably at-
tributable to their work.
Clients sometimes manipulate the gratification received by street-level
bureaucrats in order to affect future interactions. Client strategies include
passivity and acquiescence, expressions of empathy with workers' problems,
and humble acceptance of their own responsibility for their situation. The
disadvantaged position of clients forces them to conspire in their own man-
agement in order to avoid offending the workers or providing negative evi-
dence about their character. In some circumstances clients can effectively
express anger or demand their rights, but these strategies appear useful only
in certain circumstances and usually not for long. 9
While a client has some resources with which to affect a relationship with
street-level bureaucrats, the relationship is by no means a balanced one. It is
a relationship of"unidirectional" power in which "the capacity to make and
carry out decisions is the exclusive, or near exclusive, property of one of the
... groups." 10 The relationship is primarily determined by the priorities
and preferences of street-level bureaucrats, but the character and terms of
the relationship are substantially affected by the limits of the job.

The Social Construction of a Client 11

People come tQ street-level bureaucracies as unique individuals with dif-

ferent life experiences, personalities, and current circumstances. In their
encounters with bureaucracies they are transformed into clients, identifiably
located in a very small number of categories, treated as if, and treating them-
selves as if, they fit standardized definitions of units consigned to specific bu-
reaucratic slots. The processing of people into clients, assigning them to ca-
tegories for treatment by bureaucrats, and treating them in terms of those
categories, is a social process. Client characteristics do not exist outside of
the process that gives rise to them. An important part of this process is the
way people learn to treat themselves as if they were categorical entities.
If"reality" is a social construction, general agreement about its character-
istics can still be approached if the people who comprise the interaction
about the reality are not in conflict, or if they are relative equals. Spectators
at a boxing match may approach agreement on what is going on in the ring al-

59
CONDITIONS OF WORK

though they come from different class backgrounds (even here we need to be
cautious, since fighting may mean something quite different to people of dif-
ferent classes). Likewise, the boxers may have a similar view of reality be-
cause they share the common goal of putting on a show, or because, while in
conflict, they are relatively evenly matched.
There is little agreement, however, on the picture of reality where clients
and street-level bureaucrats are concerned. This is at least because the two
are intrinsically in conflict over objectives and the relationship is drastically
unequal. What street-level bureaucrats think they do may have little con-
nection with what clients think is going on. Clients tend to experience their
needs as individual problems and their demands as individual expressions of
expectations and grievances. They often expect treatment appropriate to
them as individuals, and are in large measure encouraged in this expectation
by public institutions and society in general. On the other hand, street-level
bureaucrats experience client problems as calls for categories of action. Indi-
vidual client demands are perceived as components of aggregates. Expecta-
tions of proper treatment are framed in terms of satisfactory solutions for the
optimal processing of the totality of the work rather than in terms of the best
solution for individual cases. Clients seek services and benefits; street-level
bureaucrats seek control over the process of providing them.
There are four basic dimensions to the control exercised by street-level
bureaucrats over clients. Each significantly affects some dimension of client
"construction." Briefly, street-level bureaucrats exercise control in: (1) dis-
tributing the benefits and sanctions that are supposed to be provided by the
agencies; (z) structuring the context of clients' interactions with them and
their agencies; (3) teaching clients how to behave as clients; and (4) allocating
psychological rewards and sanctions associated with clients entering into
relationships with them.

DISTRIBUTING BENEFITS AND SANCTIONS

In allocating benefits and sanctions street-level bureaucrats obviously af-
fect the relative well-being of their clients. They contribute to change and
development, to the resources clients control, and to the status clients suffer
or enjoy.
While eligibility for public service benefits often may seem cut-and-dried,
a considerable part of eligibility is in fact problematic. Rules and regulations
provide only a measure of guidance in determining eligibility. This may be
because discretion must be used in determining eligibility for the presenting
situation, as in the case of a legal services lawyer who must decide whether a
prospective client is faced with an emergency and therefore eligible for ser-

6o
 Relations with Clients
vices under agency policies. It may be because classifying the behavior or
background of the client is a matter of discretion, as in assessing whether the
behavior of a pupil or prisoner constitutes insolence. Or it may be because
the categories into which clients fit are actually problematic and not fixed, as
in judicial determination that a juvenile defendant is" a basically good kid in
trouble" rather than "a rotten apple. " 12 To the extent that the assigm;nent of
benefits or sanctions is negotiated between street-level bureaucrats and
clients through interpersonal strategies and implicit maneuvering, the allo-
cation of benefits and sanctions are clearly part of a process of constructing
the client profile.

STRUCTURING THE CONTEXT

This leads to the second dimension of street-level bureaucrats' control
over clients. The most important aspects of interactions with clients are
those affecting the structure of the interactions: when they will take place,
with what frequency, under what circumstances, with what resources com-
manded by the parties. The structure of interactions limits and determines
the range of behavioral actions from which clients may choose their respon-
ses.13 Street-level bureaucrats organize the context of decision making so
that they are able to process clients under circumstances most favorable to
controlling their behavior. In this they are not constrained by fear of client
retaliation and for the most part can impose on the clients whatever costs are
involved. Thus street-level bureaucrats develop routines that prepare peo-
ple for client status, and their agencies impose standardized ways of proces-
sing people to maximize utilization of agency resources (see chapters 7 to 9).
In this way disruptive, antagonistic, or uncooperative client behavior is dis-
couraged before it surfaces.

TEACHING THE CLIENT ROLE

A third dimension of control over clients by street-level bureaucrats con-
sists of teaching the client role. We observe the teaching of the client role in
schools, where children are socialized to the procedures of orderly
classrooms (and to the requirements of modem industrial and administrative
work). Children are taught to raise their hands if they wish to speak, walk in
twos holding a partner's hand, and organize their work to begin and finish
within set periods.
While many such lessons are appropriately treated in terms of the routines
established to process work (see chapter 9), there are several aspects of client
role instruction that may be taken up here.
Imagine a situation in which a relatively modem bureaucracy has to rou-
CONDITIONS OF WORK

tinely process clients who have no concept of bureaucratic expectations.

Heavy costs may be laid on helpless clients. The dehumanizing mass proces-
sing of immigrants at Ellis Island during the mass immigrations from south-
ern and eastern Europe is an example. But where the clients' cooperation is
necessary for the bureaucrats to function smoothly-that is, where failure to
cooperate, out of anger or ignorance, is likely to impede bureaucratic func-
tioning, the bureaucracy has a stake in teaching clients what is expected of
them. This phenomenon has been observed in the case of new immigrants to
Israel, who, coming from traditional societies, often lack simple under-
standing of the functioning of modern social institutions.
Consider, for example, the bus driver who gets out of the bus to teach the idea of a
queue-"first come, first served"-an idea which is new to many of his new im-
migrant passengers. Similarly, the nurse at the well-baby clinic may be seen teaching
women, informally, which of their needs are appropriate to the health services and
which should be taken to other organizations. 14
Teaching the client role is not confined to children or immigrants new to
modem institutions. While novices may have to be instructed from the start,
it is more commonly the case that people are attuned to the requirements of
their dependency and only incidentally reminded of proper client behavior.
Most of the time people act appropriately, serving~ models for others. But
they will be spurred to proper behavior if they stray. The uniform, night-
stick, and gun of the police officer is sufficient to cue most citizens to respect
police authority if they become inclined not to. Teachers have a variety of
small rewards (stars) and punishments (minutes after school) to help keep
children in line.
Sustained deviance from expected patterns can often have severe conse-
quences for offenders. Legal services lawyers have been observed to walk
out of interviews with clients (a very unprofessional thing to do) when clients
do not permit them to dictate the terms of the interview. 15 Educators can
virtually revoke children's right to education if they persist in disrupting
classes. Police expect citizens to obey orders. The occasional incidents in
which deaf or mentally handicapped persons are shot by police who think
they are normal but defiant citizens provide tragic testimony to the
sanctions delivered to people who fail to behave as expected.
There are at least four dimensions of the client role that street-level bu-
reaucrats convey. First, as this discussion has already suggested, they con-
vey cues as to the degree of deference expected. Proper deference will vary
from service area to service area. School psychologists may encourage stu-
dents to demonstrate independence; judges may demand absolute respect
 Relations with Clients
for symbols of the judicial process. Police expectations of gang members
have been acutely observed.
The police ell:pect law-abiding citizens to express their respect for the law by address-
ing its representatives with various gestures of deference. It is desired that the sus-
pect's physical presence communicate civility, politeness, penitence, and perhaps
fear. In addition, the use of such terms as "Sir" and "Officer" are ell:pected as indica-
tions that the humble status of juvenile is properly understood. 18
Similar observations could be made for each bureaucracy concerning the
degree and forms of respect expected by workers.
Second, street-level bureaucrats communicate the penalties for failing to
display proper deference. Teachers and policemen manifestly teach lessons
in deference by example (''I'm going to make an example of you"). They will
not easily let an affront to their authority remain unchallenged, since to do so
would be to teach the contrary lesson, that lack of deference will not be
punished. Again, lessons of this sort are usually taught subtly. Menace,
threat, or punishment will more often be hinted at than carried out.
Third, street-level bureaucrats convey to clients what their proper level of
expectations of the bureaucracy should be. This is complicated because at
one level the society generally teaches that citizens have a right to equal
treatment and responsive services. The ideology of public service urges peo-
ple to seek safety, health care, education, and other objectives from public
agencies. Diffuse public awareness that bureaucracies are characterized by
red tape, inefficiency, corruption, and bungling helps to deflate expecta-
tions, but it does not do so entirely, since individuals can still hope for and
seek out responsive treatment.
At the individual level street-level bureaucrats often convey to clients that
they should expect few services. Teachers' tracking of students, even when
done informally, precisely conveys to students that they are not expected to
learn a great deal. Police directly inform citizens that they cannot expect
much action on recovering stolen property. Welfare clients are told by social
workers that there is nothing that can be done to increase their benefits.
If nothing truly can be done, it is proper to convey this to clients. The
problem is that "nothing can be done" is only another way of saying that the
bureaucracy or individual worker does not intend to change priorities. Yet it
is often obvious to clients that more could be done if priorities were shifted.
The Welfare Rights Organization for a time successfully campaigned to win
benefits to which members were already entitled, but which had not been
forthcoming because local welfare offices had not sought to maximize client
benefits. Similarly, during the rent-strike movement in New York City in
CONDITIONS OF WORK

1g64-1965, tenants were able to gain priority with local housing agencies
because of their association with rent strike organizations. 17 "Nothing more
can be done" often really means: "Priorities will not be changed in your case,
although they could be." Since priorities could be and often are changed
frequently, bureaucracies have a stake in concealing the mutability of policy.
(It may be that bureaucratic managers often do not themselves believe or un-
derstand that priorities could be changed.)
At the agency level, bureaucracies also attempt to convey proper levels of
expectations. Long lines not only discourage prospective clients but also
convey that many people have to be processed; hence individual clients
should appreciate that workers have little time to spend with them or on
their problems. Conducting welfare intake interviews in a single intake of-
fice so that workers and clients can overhear all the interviews being con-
ducted conveys to clients that they may not expect privacy. Lower-court ar-
raignment sessions clearly convey to waiting defendants that the judge is
very busy, that the court has developed procedures to speed the process,
that the court appreciates cooperation with the process, and that most peo-
ple accept their inability to understand what is going on.
Street-level bureaucrats often attempt to involve clients in the difficulties
of their jobs in order to gain understanding or sympathy for their position.
Assertions that ''I'm just doing my job," or ''I'm following orders" help bring
the client to an agency point of view. The client is implicitly asked to aban-
don his or her own interest in the interaction in a friendly, not overtly
conflictual tone. But there is little choice involved, since the structure of the
institution requires the client to comply or else risk alienating the more pow-
erful street-level bureaucrat. No doubt many workers genuinely assert such
claims, and they are often genuinely appreciated by clients. Still, in poten-
tially conflictual stuations these claims function to gain client compliance in a
persuasive rather than coercive way.
Fourth, under some circumstances street-level bureaucrats convey infor-
mation about how to work the system. In so doing, they alert clients to the
alternatives available to them under the current structure. Most clients
would like to know more about how to negotiate the system, but this infor-
mation is rarely provided to all clients. Rather, street-level bureaucrats exer-
cise discretion by providing this information on a selective basis. This be-
comes one of the few ways they are able to favor clients without directly
abridging bureaucratic norms of fairness. They make no decisions in favor of
one client over another. They simply inform clients selectively how to utilize
the system to best advantage. Thus they respect fairness in decision making;
it is only information that is selectively distributed. Clients who are so fa-
 Relations with Clients
vored, however, receive a tremendous benefit when norms of universalistic
decision making operate. Knowing how to position oneself so as to increase
the probability of a favorable decision is a substantial advantage. 18
Sometimes teaching clients how to work the system consists of favoring
some clients by providing them with special information. Public housing
officials in Boston did this in coaching elderly applicants on how to apply
for emergency housing status. 19 Boston housing inspectors did this by re-
minding landlords that they could petition for certain hearings. 20 Sometimes
teaching the system takes the form of discriminating against some clients by
denying them information given to others. Welfare officials in North Caro-
lina apparently did this in informing most but not all clients of their right to
apply for assistance. 21
At other times street-level bureaucrats teach clients how to get best re-
sults from other bureaucracies. This practice is common among those work-
ers who prepare clients for other agencies. Thus a social worker will instruct
a client how to obtain favorable treatment in a referral. A lawyer will instruct
a client how best to impress a judge. 22
Coaching selected clients at times seems innocuous, but it almost always
has redistributive effects. The defendant instructed to appear penitent
makes others who fail to receive coaching appear defiant in contrast. The
public housing officials who coached elderly applicants on how to fit into
emergency categories helped these applicants to receive better units more
quickly, to the disadvantage of other, equally deserving applicants. Prospec-
tive welfare recipients were harmed by the failure of welfare officials to
explain their right to apply for services. 23
To the extent that biased coaching remains secret, one would hypothesize
that public legitimacy and regard for the agency would increase. No one
would perceive himself or herself to be worse off, while some would feel
favored. But to the extent that these biases become known or are exposed,
they would tend to undermine the agency's legitimacy.

PSYCHOLOGICAL BENEFITS ANV SANCTIONS

A fourth dimension of control over clients consists of producing the psy-
chological benefits and sanctions that result from client involvement with the
bureaucracy or accompany client status. There are two aspects to this psy-
chological dimension. The first concerns the rewards and penalties acquired
within the process of submitting to interaction with the bureaucracy. The
second concerns the implications for the larger society, which responds to
client status assigned by the bureaucracy.
In any interaction people see and exchange signals concerning their
CONDITIONS OF WORK

regard for each other. 24 When one person in an interaction has status and
power relative to the other, the signals emanating from that person are par-
ticularly potent. Since a person's self-concept is substantially a function of
the response of others who are important to the person, interactions with
street-level bureaucrats have psychological as well as material implications.
The psychological implications of interactions with street-level bureau-
crats may be fleeting where the interaction is not sustained. Police tend to
treat people they apprehend scornfully or respectfully depending upon their
apparent moral worthiness and the respect they display for the police. To be
treated with respect, or with utter disrespect, by these symbols of authority
have implications for citizens' views of themselves. People stopped by the
police in a sense discover whether they are or are not the kind of person to
whom respect is normally granted.
Frequently, citizens do not understand why police have stopped them or
singled them out. They attribute an officer's brusque and imperious manner
to one of their own personal or physical traits. If they are members of minor-
ity groups they may conclude that their racial or ethnic identity triggered
police intervention, confirming once again the hazards of minority status and
their belief that police officers are racist.
Anger at bureaucrats in general may be attributed to the dehumanizing
aspects of having to seek service through bureaucracy. However, for most
people apprehension by the police, visits to emergency rooms, or court ap-
pearances are single events which have limited implications for personality
development. But any of these events may lead to further bureaucratic en-
tanglement, may reinforce patterns of interaction previously encountered,
or may signal other bureaucracies that the same (damaging or rewarding) ori-
entations toward the client should be forthcoming in the future.
The greater the involvement of the client with the agencies and their em-
ployees, the more sustained and critical the psychological implications of the
interactions. For this reason the dependence of poor people on government
services creates the context in which interactions with street-level bureau-
crats may have substantial psychological implications. At the very least poor
people who bounce from one agency to another have reinforced feelings of
dependency, powerlessness, and, deriving from these, anger. After sus-
tained exposure to the welfare system, for example, recipients have been
found to see themselves as "undeserving" and "lucky to get anything at
all. "25
Institutions that fully dominate peoples' lives have extensive influence
over personality development. As Erving Coffman has demonstrated, men-
tal hospitals teach patients how to be patients by rewarding behavior that

66
 Relations with Clients
conforms to staff expectations of how mentally ill people behave. Thus they
not only teach the client role but touch the person playing the role as well,
since for mental patients the role is also their salient personal identity. 26
The closer institutions get to total involvement with clients, the more
their self-images may be affected in a sustained way. Therapists and counse-
lors utilize this fact to reinforce positively the self-images of their clients.
Through residency in supportive halfway houses, peer support groups, and
other environmental influences, some therapists seek not only to help clients
develop psychologically, but also to insulate them from contradictory imag-
ery during the period of involvement. 27
For most people, the street-level bureaucrats whose psychological reac-
tions are most powerful are teachers. Teachers powerfully convey images to
children concerning their expectations of achievement. These images affect
the child's self-image, self-expectations, and actual achievement. Through
formal, and more insidiously, through informal tracking, teachers indicate to
students who is expected to achieve and who is not.
In terms of the outcomes of institutional involvement with citizens, the
policy problem is not only that students labeled as likely or unlikely to
achieve do perform at expected levels, thereby fulfilling the original achieve-
ment prophecy. 28 It is also that determinations concerning the likelihood of
success are not even made on the basis of achievement potential, measured
however crudely, but are made on the basis of inferred social class, or the
biased "harder" data accumulated or reported by other street-level
bureaucrats.
In a careful study of a group of primary school pupils, Ray C. Rist ob-
served that kindergarten pupils were placed in putative ability groupings
solely on the basis of their dress, demeanor, verbal skills, and social back-
ground, before any sustained interaction between the teacher and the chil-
dren and without testing. 29 The fast-track children received most of the
teacher's instructional attention and most of the classroom rewards, although
performing honorific chores did not require cognitive skills. The group
placed in the lower classroom tracks learned from the teacher that they were
not expected to achieve and that they were the kinds of children who could
be ignored and ridiculed by authoritative adults (the teacher) and their peers
with higher alleged learning ability (the children in the fast track). Mean-
while, the fast-track students were instructed in :everal ways about their
proper self-image. They learned not only that they were brighter than the
other children, but also that it was legitimate to scorn the others.
Rist noted that in subsequent grades the relatively poor performances of
the slow-track children (perceived as such by the teacher, who had insured
CONDITIONS OF WORK

that it was so) were reified into fact and formed the basis for other teachers'
tracking placements. He also found indications that the lower-track children
were learning the material, despite the fact that the teacher tended to ignore
them. The teacher did not know this because she did not encourage the
lower-track (and lower-class) children to display their knowledge, and the
cumulative effects of the classroom tracking system discouraged them from
displaying their knowledge.
Analogous lessons are communicated in interactions in other street-level
bureaucracies. Their potency depends on the extent to which clients have
similar dependency and sustained interactions. The reification of street-level
bureaucrats' prior judgments in subsequent placement decisions is also com-
mon. Judges' decisions concerning severity in sentencing juveniles, for ex-
ample, depends substantially upon the defendant's record rather than on the
severity of the offense. 30 Disposition of students with behavioral disabilities
also depends on the previous incidents recorded. 31 In both these instances it
is easy to see that young people are subject to the hazard that in the past they
may have encountered adults who regarded their behavior as requiring of-
ficial action, while others, perhaps because of differences in class back-
grounds or in the political culture of the institution, did not have offenses
recorded.
The example of the informal tracking of primary school children directs at-
tention to the second dimension of labeling by street-level bureau-
cracies-the implications of client status for the larger society. Not only do
slow-track children incorporate in their images of themselves the perspec-
tives of the teacher and the fast-track pupils. The label of"slow learners" also
has relevance for other people who play significant roles in the childrens'
lives. Parents learn that their children have been assigned to the slow track
and may begin to treat them as dull or with anxiety that they have been con-
signed early to failure. Brothers and sisters may tease. As Rist reports,
teachers in later years will treat the children as having the capacities as-
signed originally by the first teacher. There is little escape from the implica-
tions of this labeling. Sometimes people can alter stereotypical images of
themselves by presenting massive evidence to the contrary, 32 but in this
case the process of interaction is structured so that all the evidence seems to
confirm the original diagnosis.
Differences attributed to clients may be regarded as significant when peo-
ple important to the client respond to him or her differently. When dif-
ferences reach this level, a person's self-image and self-respect are affec-
ted.33 The status of "criminal," "juvenile delinquent," "welfare mother,"
and "slow learner" is stigmatic because it goes beyond mere distinctions

68
 Relations with Clients
among people. Society takes these terms as signals to treat people dif-
ferently. As such, they are simultaneously cues to the people so labeled to
regard themselves differently.
Some differences assigned by street-level bureaucrats provide more sub-
tle stigma. For example, being a public housing tenant is not necessarily
stigmatic, but living in certain projects may be. Health treatment through
medicare may not affect a client's relationship with some clinics, but might
have substantial implications in others. The label of "troublemaker" may
predict official responses in some schools but not in others.
Prisoners are never fully destigmatized by the society after they have
served time. They become "ex-cons." Indeed, mere involvement with a bu-
reaucracy may be stigmatizing, even when the clearly stigmatic label is
avoided. This may be the fate of a defendant found innocent at trial, 34 or the
patient who is held for examination to determine whether he or she requires
treatment for mental illness, even if judged perfectly sane. If others begin to
treat one differently as a result of these labels, one may begin to incorporate
these views as his or her own self-image.
The particular difficulty with labels ascribed by street-level bureaucrats,
as we have seen, is that the characteristics on which they are based are prob-
lematic. Judgments concerning the status allocated by street-level bureau-
crats depend upon the discretion of the bureaucrat, which in turn depends
upon many indeterminate factors, such as training, the social context in
which the client is presented, and the presence or absence of similar "dif-
ferences" in the client population. The fast-track students in the above illus-
tration might have been labeled slow learners in white, middle-class schools
(the students and teachers were black). A criminal offense in one setting
might be overlooked in another. The social construction of the client, involv-
ing the client, others relevant to the client, and the public employees with
whom they must deal is a significant process of social definition often unre-
lated to objective factors and therefore open to the influences of prejudice,
stereotype, and ignorance as a basis for determinations.
Some bureaucracies so routinize their processing of clients that significant
psychological interactions are minimal. Welfare workers and legal service
lawyers, for example, may adhere to interview formats that exclude personal
elements and reduce the likelihood of decision making on the basis of inter-
personal interactions. 35 Of particular interest to those concerned with the
policy consequences of street-level bureaucrats' work is the ways in which
the tendencies to treat people in terms of their predicted behavioral charac-
teristics can be avoided. Some bureaucratic settings seem to result in stigma-
tic treatment and some do not. There are a variety of ways to interrupt track-
6g
CONDITIONS OF WORK

ing patterns without abandoning educational objectives. Over time,

recognition of the potential of people previously defined as physically and
behaviorally deviant-as in the example of exceptional children-can
change. These are aspects of the policy-making roles of street-level bureau-
crats that can be influenced only through _analyzing the interaction between
those who assign status and those who are assigned, and affecting the work
context in which the social construction of the client takes place.
 CHAPTER 6

Advocacy and Alienation zn

Street-Level Work

To deliver street-level policy through bureaucracy is to embrace a contra-

diction. On the one hand, service is delivered by people to people, invoking
a model ofhuman interaction, caring, and responsibility. On the other hand,
service is delivered through a bureaucracy, invoking a model of detachment
and equal treatment under conditions of resource limitations and con-
straints, making care and responsibility conditional.
The human model of interaction contributes to the motivation of public
service workers, who believe they are helping others, and to the motivation
of clients, who are encouraged to confide and trust in strangers and permit
themselves to be manipulated and ordered about in the expectation of re-
ceiving help or fair treatment. The foundation of support for the human
model of interaction rests in the belief that this model benefits clients. Yet it
may be called a myth of altruism because the assertion that agencies provide
benefits and fair treatment is usually unexamined, not subject to falsification
among people who believe it, and a means for structuring a range of further
assumptions about public policy . 1
Whatever the precise functions of this myth, the importance of promoting
it may be judged by the intensity with which it is perpetuated. It is perpetu-
ated in professional canons of ethics, which instruct professionals to treat the
whole person, to respect and encourage client autonomy, and to respond to
the individual rather than to alleged patterns of group behavior. It is perpet-
uated in schools of training, whose curricula teach bodies of knowledge (for
example, the law, or educational psychology) or, rarely, how to treat individ-
CONDITIONS OF WORK

uals (for example, interviewing techniques), but virtually never how to jug-
gle case loads or handle large numbers of clients at one time, as initiates will
have to do. One of the best illustrations of the solidity of the myth of human
interaction in public services is provided by the transformation in the health
field of the word "care" from a verb to a noun. Politicians and administrators
regularly discuss levels and amounts of care that will be provided, but rarely
who will care, and how they will express their caring.

Advocacy

Street-level bureaucrats are often expected to be more than benign and pas-
sive gatekeepers. They are also expected to be advocates, that is, to use their
knowledge, skill, and position to secure for clients the best treatment or
position consistent with the constraints of the service. That street-level bu-
reaucrats should be advocates for clients is articulated explicitly in the pro-
fessional training and canons of lawyers, doctors, social workers, teachers,
and others. Those professions and semi-professions that display the altruism
critical to most definitions of professionalism require their members to make
clients' needs primary. Other street-level bureaucracies whose claims to
professional status are more questionable also display degrees of advocacy in
their obligations as public servants to be responsive to the citizens who pay
their salaries.
One source of the myth of service altruism is social policy reformers who
utilize the discrepancy between reality and stated policy intentions to mobi-
lize support for change. Virtually an policy reforms are advocated in the
name of achieving service ideals. (The myth of altruism does not assume
ideal policy implementation; it assumes only that policy and people who
implement it are well intentioned and that their work constitutes a net social
benefit.)
Perhaps even more important in sustaining the myth of service altruism
are the workers who attempt to implement the service ideal. Like most
social myths this one has a partial basis in reality. Each generation of workers
brings to its jobs, in addition to interest in material benefits, dedication to
helping people. Those who recruit themselves for public service work are at-
tracted to some degree by the prospect that their lives will gain meaning
through helping others.
The myth of human relationships in street-level bureaucracies no doubt

72
 Advocacy and Alienation in Street-Level Work
partly accounts for whatever dedication exists to superior service and for
the ability of public services to realize their objectives. However, achieve-
ment of advocacy is undermined by several critical factors. Some of these
concerns were treated earlier in discussing the structure of street-level bu-
reaucrats' work, but it is useful to review briefly how the structure of work
and relations with clients compromise altruism and undermine advocacy
where advocacy is appropriate.
The helping orientation of street-level bureaucrats is incompatible with
their need to judge and control clients for bureaucratic purposes. This is evi-
dent in the following role tensions.
First, advocacy can only be done on behalf of single units, whether they
be individuals or collectivities such as a tenants' union. Moreover, the ad-
vocate must have enough free attention to devote to the client. This does not
mean that only one client can be dealt with at a time. But it does mean that
advocacy may be compromised by large case loads and mass processing of
clients. For the advocate, large case loads mean that every minute devoted
to one client means less time for others. Clearly organizations have to choose
what resources to provide, and a suboptimal amount is likely to be available
for any client. Street-level bureaucracies chronically tend to allocate rela-
tively low amounts of resources to facilitate workers taking clients' perspec-
tives. Clients are asked to understand this and even to incorporate this un-
derstanding into their concept of being a client. Those who fail to do so are
punished by bureaucracies, which can impose severe costs on clients who
contest the allocation of resources devoted to them.
Second, advocacy is incompatible with organizational perspectives. The
organization hoards resources; the advocate seeks their dispersal to clients.
The organization imposes tight control over resource dispersal if it can; the
advocate seeks to utilize loopholes and discret-ionary provisions to gain client
benefits. The organization seeks to treat all clients equally and to avoid hav-
ing to respond to claims that others received special treatment; the advocate
seeks to secure special treatment for individual clients. The organization acts
as if available resource categories had fixed limits (which is often not abso-
lutely true); the advocate acts as if resources were limitless (which is also not
true).
Street-level bureaucrats frequently encounter this tension. School coun-
selors are criticized for serving the interests of schools rather than individual
children. 2 Doctors have come under considerable criticism for failing to give
sufficient priority to not spending public monies for patients' health care.
This is another way of saying that their advocacy perspective (advocacy for
their patients and, perhaps, themselves) is not in balance with their cor-

73
CONDITIONS OF WORK

porate responsibilities. 3 Potentially the same situation exists anytime street-

level bureaucrats provide services with open entitlement. For example, the
1972 special education law in Massachusetts directed that educational plans
for special needs children be developed and implemented irrespective of
costs. State funding for this open-ended provision was not forthcoming, how-
ever, and local school systems were required to spend an apparently limit-
less amount of previously unauthorized funds. In practice these boards did
take school system resources into account despite specific provisions to the
contrary in the law. 4
Third, advocacy is incompatible with controlling clients. Street-level bu-
reaucrats usually must make judgments about clients on matters unrelated to
appropriate service. They must as well make judgments about credibility, eli-
gibility, and performance. Is the welfare recipient truthful? Does the claim-
ant of legal services truly face an emergency? Is the job trainee certifiably
competent? The street-level bureaucrat is almost always a judge as well as a
server. Yet it is hard to do both at the same time. Since these are human in-
teractions that are the subjects of judgment, street-level bureaucrats are not
free to give themselves unreservedly to clients. They feel the need to make
sure that they do not lose control, respect, advantage, or face, or otherwise
fail to perform as required by their role. Street-level bureaucrats may at-
tempt to do a good job, but it will be a job tempered by the other psycho-
logical and role requirements placed upon them. 5
Fourth, advocacy is incompatible with the responsibility of street-level
bureaucrats to prepare clients for presentation to other workers or other bu-
reaucracies. One of the most substantial checks on workers who deal with
clients is the social and other pressures that arise from the fact that a client is
later seen and processed by still other workers or is presented to outsiders.
To be sure there are norms against peer criticism in some areas, such as
medicine, but pressures exerted by the anticipation that others will observe
the work are nonetheless substantial. Examples include police officers who
anticipate presenting cases in court, public defenders who seek to control
clients so as not to embarrass themselves before judges, 6 and employment
counselors who try to maintain credibility with employers by recommending
only the best candidates for jobs. 7
Several of these dilemmas are presented by the example of the social
workers in Washington state assigned to child protective services. Individ-
uals in these jobs are empowered to investigate and provide authoritative ad-
vice to judges in hearings concerning the removal of children from their
parents' custody. Yet they are also supposed to provide counseling for
parents, primarily from the perspective of serving as an advocate for the

74
 Advocacy and Alienation in Street-Level Work
child. In this four-sided relationship (agency, judge, parents, child) social
workers might well yearn for the clarity an entirely adversarial process might
provide. 8

Alienation

There is another perspective that illuminates the relative attractiveness of

street-level positions. Street-level bureaucrats' work is alienated work. This
alienation not only affects their commitment to jobs and clients, but, as it
also affects the quality of their vocational experiences, it is a significant state-
ment about public policy itself, considering the millions of people who are
engaged in street-level employment.
The promiscuous word "alienation" has been overused and under-
appreciated. I am not interested here in alienation as a psychological orienta-
tion, although psychological orientations no doubt arise from alienated
work. 9 Rather, alienation as used here is a concept summarizing the rela-
tionship of workers to their work, from which, we may infer, attitudes arise.
Worker alienation summarizes several concerns: the extent to which the
worker makes decisions about the work, has control over what is made and
how it is fashioned, and influences the disposition of the product. Alienation
at times also refers to the extent to which workers are able to express, or
need to suppress, their creative and human impulses through work activity.
Assembly lines, in which mechanical, repetitive processes are performed,
are regarded as alienating for these reasons. 10 Jobs that require workers to
deny the basic humanity of others may also be considered alienating.
I have already indicated how some of the shared working conditions of
street-level bureaucrats appear to be characteristically unalienated. For ex-
ample, discretion about clients separates street-level bureaucrats from as-
sembly line workers in two respects. Bureaucrats make decisions about the
product of their work and they work with (or on) people, so that they are con-
stantly confronted with the variety ofhumanity. (For po1ice officers the most
rewarding aspect of an otherwise often unrewarding job is the variety of
situations and people they encounter.) To the extent that the pressures of
the job do not entirely narrow these opportunities, aspects of street-level bu-
reaucrats' work contribute to their fealty toward the diffuse objectives of
their agencies.
Street-level bureaucrats often have relatively good relations with other
workers. The peer structure in street-level bureaucracies often is quite

75
CONDITIONS OF WORK

strong. Street-level bureaucrats work in isolation, but they seek and receive
support from other workers. Street-level bureaucrats are no less peer-
related than other workers and find gratification in the squad room, teachers
lounge, and other places where they congregate.
However, there are several areas in which the alienation of street-level
work is fairly extensive. I have already said that the compromises required of
advocates reduce the extent to which street-level bureaucrats are able to
respond to clients in a fully human way. Moreover, street-level work is
inauthentic, in the particular sense of alienation delineated by sociologist
Amitai Etzioni. In its emphasis on providing services, mandating workers to
act as helpers, and giving them responsibility it "provides the appearance of
responsiveness while the underlying conditions . . . subject a person to
forces beyond his understanding and control." 11 In defense of the myth of al-
truism, street-level bureaucracies devote a relatively high proportion of
energies to concealing lack of service and generating appearances of
responsiveness.
In addition, street-level bureaucrats are alienated from clients-the prod-
uct of their work-in at least four particular respects: (1) they tend to work
only on segments of the product of their work; (2) they do not control the
outcome of their work; (3) they do not control the raw materials of their
work; and (4) they do not control the pace oftheirwork. 12 These considerations
are discussed below.
1. Working on segments of the product. One difference between shoe-
makers and shoe-factory workers is that the former crafts the entire shoe and
thus can draw satisfaction from seeing the fruits of his or her labor. The latter
may only cut the heel, punch holes for laces, glue parts together, or perform
some other aspect of the shoe making process. But he or she cannot take pride
in having made the article. This analysis of the implications of factory work is
as old as the factorization of production and as young as recent industrial ef-
forts to permit workers greater flexibility and variety in work assignments. 13
Street-level bureaucrats do not work on the entire product, but only on
segments. This is the case in two respects. In response to the need to cat-
egorize clients they tend to treat them only as bttndles of bureaucratically
relevant attributes rather than as whole persons. They deal with symptoms,
qualifications, and capacities, but not with feelings or superficially tangential
facts. The imperatives of processing people into the correct categories tend
to overwhelm both professional obligations to treat the whole person and the
recognition that responding to clients in narrowly defined areas is likely to
miss important dimensions of the presenting problem.
 Advocacy and Alienation in Street-Level Work
Understanding that treating parts of people leads to inferior or inappro-
priate services has generally been the guiding idea behind many critiques of
social policy. For example, social workers and psychologists have been em-
ployed in hospitals, schools, and courts in efforts to respond holistically to
the citizen-client. Reformers recognize that a health problem such as lead
poisoning may be a problem of income and law. A problem of poverty may in
reality be a legal problem (e.g., being granted a divorce). Education prob-
lems may have emotional, physical, or economic origins. The more the soci-
ety recognizes the interconnectedness of service policy problems, and the
more the problems are or remain interconnected, the more the alienation of
categorization will impinge on street-level work.
Street-level bureaucrats also tend to work only on segments of the pro-
cess. In the name of efficiency, convenience, or optimal utilization of re-
sources the world of social services has become n:tore and more specialized.
Educators are math teachers, reading teachers, art specialists, dance thera-
pists. The world of special education now includes therapists on various
forms of reading disability, various degrees of mental retardation, and
various kinds of physical handicaps.
It would be one thing if these specialists had relatively intensive interac-
tions with children so that they could fully plan for them and have the time
to realize measurable results. But school programs tend to be filled with spe-
cialists who cannot take full responsibility for the product even if they
wanted to. This is not an argument against mainstreaming, for no doubt
children with special educational needs ought not to be segregated. But
mainstreaming in the modern school means going into a system that substan-
tially segments the child's school day. Schools not only specialize by func-
tion, but they also are organized so that different workers take responsi-
bilities for different stages through which pupils pass. Different teachers
receive students each year as teaching is cross-specialized by grade. The ex-
ample of specialization in teaching is particularly apt because schools have
the most extensive opportunities to interact substantially with service
recipients.
There are often considerable costs to specialization in addition to the
benefits of expertise and efficiency. Divisions between intake and casework
mean that interviews and fact gathering sometimes have to be repeated,
which is inefficient. When resource allocation does not permit extensive
reinterviewing of clients the result may be inappropriate decisions resulting
from the distortion of information between intake and case-work levels.
Recent proposals for multi-specialization in special education appear to pro-

77
CONDITIONS OF WORK

vide support for the view that clients of special education services have many
dimensions, and that treating the whole person is required ultimately for op-
timal results.
z. Controlling the outcome of the work. For reasons closely related to the
above, street-level bureaucrats often do not control the outcome of their
work. Specialization may mean that they do not see the work through or only
participate in a fraction of the work with clients. They do not control all of
the resources of the agency they work for. Sometimes they process people
for other bureaucracies, which ultimately disposes of cases. Police often
conplain, for example, that their actions are not supported by judges and
prosecutors; they feel uncertain whether their performance of a good job as
they define it will result in a desired outcome.
Another reason they do not control the outcome of their work is that
clients' problems are not subject to closure. Although street-level bureau-
crats are regarded, and regard themselves, as able to solve problems, the
problems do not end or are not resolvable. Many public service agencies are
called revolving doors for this reason. Street-level bureaucracies that are
oriented toward transforming clients, such as judicial institutions and social
welfare agencies, are revolving doors because the solutions they offer people
are not adequate. People do not stay "fixed." To the extent that this presents
street-level bureaucrats with severe dissonance between objectives and ca-
pabilities, they develop coping mechanisms to shield them from the implica-
tions of the gap between expectations and accomplishment. They are alien-
ated to the extent they experience this discrepancy as loss of control over
situations they are supposed to control.
3 Controlling the input. Street-level bureaucrats cannot control the na-
ture of the material with which they work. They cannot deploy to greatest ef-
fectiveness the skills they possess partly because the conditions of work pro-
hibit effective interaction with clients, and because they do not have control
over clients' circumstances even when conditions are favorable for interven-
tion. Even in institutions such as prisons and mental hospitals workers do
not control the under life of the institutions and cannot affect those realities of
clients' lives that contribute to their deviant behavior. How frustrating to be
a good teacher who has to greet every morning children who are hungry and
exhausted from lack of sleep. How frustrating to be a skilled professional at-
tempting to help a welfare recipient whose circumstances conspire to con-
found every constructive move. It takes a great deal of commitment or
cynicism to accept the futility of one's own efforts in such circumstances. 14
4 Controlling the work pace. Street-level bureaucrats do not control the
pace of work. I have said that their discretion provides a measure of reward
 Advocacy and Alienation in Street-Level Work
on the job, but they often do not control the timing of their decision making.
This is the case in obviously reactive public service areas such as policing,
and in other areas where the amount of time spent on individual clients, or
the number of clients requiring attention, cannot be anticipated. Street-
level bureaucracies go to great lengths to produce predictability of client
demands, either by rationing services in some way or planning for peak
period workloads. To some extent they are able to develop devices to mini-
mize the costs of unpredictability. (This is treated in chapter 7.) But to some
extent they are not. Then they confront the problem that their work is their
master. They get behind in paperwork, are unable to respond to any but the
most minimal requests, and believe themselves to be ineffective because
they feel they have no slack time to respond fully to any individual situation.
Alienated work leads to dissatisfaction with the job. Job dissatisfaction af-
fects commitment to clients and to the agencies for which they work. The
proposition that street-level bureaucrats perform in alienated labor roles
contributes to understanding the dynamics of some recent developments in
public service organization.

Implications of Alienation

Public service work, particularly in street-level bureaucracies, has be-

come more bureaucratic over time. Civil service and other developments
have made worker recruitment more universalistic. This development may
also have affected the extent to which workers are able to help clients and
see the results of their efforts because treatment of clients has also become
more universalistic. Congruity between the social background of workers
and clients does not have the same place in public service interactions as it
used to. The fabled paternalism of social workers and school teachers in the
past may have functioned to give street-level bureaucrats a sense of responsi-
bility and reward for outcomes, even if these gratifications were founded on
inequalities and favoritism. Thus the bureaucratization of public service may
have been accompanied by increased worker alienation . 111
To the extent that street-level bureaucrats are alienated in their work,
they will be more willing to accept organizational restructuring and less con-
cerned with protecting clients' interests and their own connection with
clients. The more tenuous the relationship with clients, the less salient that
relationship becomes, and the easier it is to transform the relationship fur-

79
CONDITIONS OF WORK

ther. Thus the working conditions that give rise to alienation in work may
cumulatively contribute to separating the client from the public service
worker. This is significant since in earlier periods public service workers
have often championed client rights and benefits. There have been times in
American labor history when fledgling public service unions and worker as-
sociations bargained for clients as well as for themselves (social workers and
teachers, for example). This struggle has become less important as the con-
nection between workers and clients has dissipated.
In general, employers confronted by alienated workers can choose among
the following responses in some combination. They can tgnore the situation
and accept the absenteeism, low morale, poor performance, and other mani-
festations of worker dissatisfaction. They can restructure the work to make it
less alienating. Or they can concentrate on changing the mix of benefits and
sanctions that they offer workers outside of working conditions. They can
increase pay for work that is less rewarding or improve conditions tangential
to the work. Or they can raise the costs of nonproductive behavior.
The trends in public employment since, say, 1g6o reflect these alterna-
tives. Public service workers have increased their share of national wealth
through higher pay and benefit levels, increased their collective bargaining
power, and acquiesced in and often encouraged developments such as spe-
cialization, computerization, and fragmentation of responsibilities for clients.
Street-level bureaucrats have enhanced their position in the political system
to the neglect of aspects of service consistent with more humanistic models
of client involvement, or at the expense of taking positions on clients' behalf.
These statements no doubt require qualification for particular service areas,
but at a general level they suggest linkages between the quality of client ser-
vices, the structure of street-level bureaucrats' work, and the priorities of a
society oriented toward cost effectiveness and the dispensary model of pro-
viding for human needs.

8o
 PART III

PATTERNS OF
PRACTICE
Introduction

THE "PROBLEM" OF STREET-LEVEL BUREAUCRACY 1

Street-level bureaucrats work with inadequate resources in circumstances
where the demand will always increase to meet the supply of services. Thus
they can never be free from the implications of significant constraints.
Within these constraints they have broad discretion with respect to the
utilization of resources (by definition). In the application of resources to the
job they confront the uncertainty that stems from the conflicting or ambigu-
ous goals that unevenly guide their work. They also confront the additional
uncertainties that arise from difficulties in measuring and evaluating work
performances. A final salient condition of work is that the people with whom
street-level bureaucrats regularly interact are not among their primary refer-
ence groups, affecting the degree to which client satisfaction has priority.
While many aspects of their work promote a client orientation, still others
lead to reducing commitment to the work. Thus the portrait of the street-
level bureaucrat is one of considerable responsibility in allocating social
values but little effective external determination as to how to define and
achieve objectives.
To this portrait of work conditions an additional consideration must be
added. Street-level bureaucrats manifestly attempt to do a good job in some
way, given the resources at hand and the general guidance provided by the
system as outlined above. 2 Street-level bureaucrats share with others the
need to think of themselves in a reasonably favorable light. Most street-level
bureaucrats can be taken at face value when they assert that they are doing
what they think is the best they can do. Typically, they do not claim that
PATTER NS OF PRACT ICE

they are doing a perfect job or performing the way the job should be per-
formed; only that they are functioning effectively and properly under the
constraints they encounter. The typical teacher, policeman, welfare
worker-ind eed anyone who regularly meets the public-seem s to have an
image of himself or herself as working under great strain and with consider-
able sacrifice to provide clients protection or service no one else would be
willing to provide. They see themselves as fighting on the front line of local
conflict with little support and less appreciation by a general public whose
dirty work they do. 3
If they have any recognition that their performance is less than adequate
under the circumstance s they confront, they are likely to seek and find the
explanation someplace other than in their own inadequacy. Street-level bu-
reaucrats who are unable to retain a concept of their own adequacy in the job
are more likely to leave it or seek other work than to sustain the personal am-
bivalence that results. This orientation applies even though the civil service
system may deaden motivation, and some public employees may be mo-
tivated by primarily selfish consideratio ns.
The point that they try to do a good job in some way fulfills the require-
ments for asserting that the problem of street-level bureaucrats is one of
decision making under conditions of considerable uncertainty where satisfac-
tory decisions about resource allocation must be personally as well as organi-
zationally derived. The work context of street-level bureaucrats calls for the
developmen t of mechanisms to provide satisfactory services in a context
where the quality, quantity, and specific objectives of service remain (within
broad limits) to be defined.
There is by now a venerable tradition in organization al studies concerning
the search for satisfactory rather than optimal solutions to decision-ma king
problems under conditions of uncertainty. 4 But the analysis of street-level
bureaucracy may be somewhat different from other studies because it is not
only the decisions that become satisfactory rather than optimal, but also the
mental and organization al processes that must become satisfactory. Thus to
understand street-level bureaucracy one must study the routines and subjec-
tive responses street-level bureaucrats develop in order to cope with the dif-
ficulties and ambiguities of their jobs. 5
We can now restate the problem of street-level bureaucracy as follows.
Street-level bureaucrats attempt to do a good job in some way. The job,
however, is in a sense impossible to do in ideal terms. How is the job to be
accomplishe d with inadequate resources, few controls, indetermina te objec-
tives, and discouraging circumstanc es?
There are three general responses that street-level bureaucrats develop to
 Introduction
deal with this indeterminacy. First, they develop patterns of practice that
tend to limit demand, maximize the utilization of available resources, and
obtain client compliance over and above the procedures developed by their
agencies. They organize their work to derive a solution within the resource
constraints they encounter. Second, they modify their concept of their jobs,
so as to lower or otherwise restrict their objectives and thus reduce the gap
between available resources and achieving objectives. Third, they modify
their concept of the raw materials with which they work-their clients-so
as to make more acceptable the gap between accomplishments and objec-
tives. Much of the patterned behavior of street-level bureaucrats, and many
of their characteristic subjective orientations, may be understood as re-
sponses to the street-level bureaucracy problem.

ROUTINES AND SIMPLIFICATIONS

In everyday life people seek to simplify their tasks and narrow their range
of perceptions in order to process the information they receive and develop
responses to it. They create routines to make tasks manageable. They men-
tally simplify the objects of perception to reduce the complexity of evalua-
tion. They structure their environments to make tasks and perceptions more
familiar, less unique. Routines and simplifications aid the management of
complexity; environmental structuring limits the complexity to be
managed. 6
Bureaucrats develop routines to deal with the complexity of work tasks.
Indeed, for some analysts routinization is virtually equivalent to bureau-
cratization. 7 For others, routinization inevitably occurs in bureaucracies be-
cause of the scarcity of resources relative to the demands made upon them. 8
The development of simplifications, as mental routinization, predictably
characterizes bureaucrats whose work involves processing the objects of bu-
reaucratic attention. At the organizational level bureaucracies officially rec-
ognize simplifying cues, such as eligibility requirements, in order to regular-
ize decision processes. 9 However, bureaucrats also develop their own
patterns of simplification when the official categories prove inadequate for
expeditious work processing, or if they significantly contradict their
preferences.
The fact that bureaucracies develop routines and simplifications is hardly
cause for comment in itself. However, the structure of these routines and
simplifications, and the structuring of the context in which they take place,
are worth considerable discussion. Where policy consists of the accretion of
many low-level decisions the routines and categories developed for pro-
cessing those decisions effectively determine policy within the parameters
PATTERN S OF PRACTIC E

established by authorities. 10 In this sense, as observed earlier, street-level

bureaucrats "make" policy.
To put it another way, the routines, simplifications, and low-level deci-
sion-making environments of street-level bureaucracies are political. Street-
level bureaucrats, as I have been arguing, determine the allocation of partic-
ular goods and services in the society, utilizing positions of public authority.
To say that their actions are political is to indicate that some people are
aided, some are harmed, by the dominant patterns of decision making. If the
dominant patterns of decision making are characterized by routinization and
simplification, then the structure of these patterns must be analyzed to de-
termine who gets what, when, and how from this sector of government. 11
The political significance of routines is highlighted by the fact that the
policies that result from routine treatment are often biased in ways unin-
tended by the agencies whose policies are being implemented or are anti-
thetical to some of their objectives. For example, the declared policy of the
life insurance industry to provide funding for black ghetto enrichment in the
late 1g6os was undermined by routine processing ofloan applications, which
tended to follow previously operative procedures. These procedures tended
to aid those individuals who were most able to receive credit anyway, to the
neglect of the originally targeted population. 12 Draft boards operating under
general regulations established by the Selective Service System routinely
implemented policies that tended to favor men from relatively affiuent fami-
lies at the expense of those from working-class families. 13 Routines in these
instances did not merely facilitate work; they determined outcomes di-
vergent from the stated policy objectives.
At times street-level bureaucrats' routines and simplifications virtually are
the policies to be delivered. Police routines established to approach motor-
ists who may prove uncooperative are immediately received as an episode
in community relations. Similarly, teachers' informal classification of pupils
by attributed learning ability effectively determines school stratification
policies.
These illustrations have special significance for three reasons. First, they
are routines of interaction. Thus citizens may be expected to react-with
subsequent implications for worker-client relationships. It is one thing to say
that bureaucracies routinize the processing of work, but when the work con-
sists of decisions made about people during the interaction itself, the sub-
jects of routinization will be affected by the processing. This clearly dis-
tinguishes analysis of people-processi ng bureaucracies from analyses of other
bureaucratic settings.
Second, while we may generally anticipate routinization and simplification
 Introduction
in human affairs, the degrees of organizational routinization and simplifica-
tion are not predetermined. At least theoretically there is a considerable dif-
ference between routinization necessary for minimally efficient functioning
and maximum routinization. Moreover, organizations can decide to be less
efficient in order also to be less routinized in their client interactions. In-
deed, routinization may prove dysfunctional at some point, complicating ef-
ficient operations. 14 Similarly, bureaucrats may be expected to categorize
clients, but the extent to which they are open to fresh information contra-
dicting facile categorization also is not predetermined.
This is particularly important for street-level bureaucrats who have a pub-
lic trust to make significant decisions about citizens' welfare. Police, judges,
teachers, and mental health workers, for example, are generally obliged to
make decisions on the basis of the available evidence rather than presump-
tions of proper determinations. They are obliged because they have been as-
signed profound responsibilities concerning the liberty of citizens or the fate
of people regarded as incompetent and unable to act in their own best
interests.
Third, routines and simplifications are subject to biases from a variety of
sources. While they often may be oriented toward fulfilling agency objec-
tives, these measures are also structured to aid workers' job requirements,
which may conflict with agency demands. Furthermore, routines and simpli-
fications are subject to workers' occupational and personal biases, including
the prejudices that blatantly and subtly permeate the society. The biases
expressed in street-level work may be expected to be manifested in propor-
tion to the freedom workers have in defining their work life and the slack in
effective controls to suppress those biases. Since street-level bureaucrats
have wide discretion about clients, are usually free from direct observation
by supervisors or the general public, and are not much affected by client
preferences, their routines and simplifications deserve considerable scru-
tiny. Sociologist Julius Roth, in introducing his study of client treatment in
hospital emergency rooms, states this perspeetive sharply.
There is no evidence that professional training succeeds in creating a universalistic
moral neutrality .... On the contrary, we are on much safer ground to assume that
those engaged in dispensing professional services (or any other services) will apply
the evaluations of social worth common to their culture and will modify their services
with respect to those evaluations unless discouraged from doing so by the organiza-
tional arrangements under which they work. 15

The analysis of routines developed by street-level bureaucrats must begin

with the proposition that they tend to contribute to control over the work en-
vironment. This is consistent with perceiving routines as coping behaviors in
ss
PATTERNS OF PRACTICE

which the confronting problem is the management of work stresses. The ten-
dency to seek control over the work environment is perhaps evident in most
work situations, but again it is worthy of comment because in street-level bu-
reaucracies the search is typically a matter of public policy. Routines could
be structured to maximize the achievement of agency objectives. Or they
could be structured to maximize responsiveness to clients. No doubt these
competing perspectives do account for workers' routines to some degree.
However, the extent to which routines are structured to maximize worker
control over the work context may measure the extent to which articulated
agency policy objectives are difficult to achieve.
The routines of work in street-level bureaucracies appear to be directed
toward achieving one or more of four purposes in processing clients.
1. They ration services.
2. They control clients and reduce the consequences of uncertainty.
3 They husband worker resources.
4 They manage the consequences of routine practice.

At times routines and simplifications will be entirely informal and contrary

to agency policy. At other times they will be consistent with agency policy
and may even be promoted by the agency. It is necessary to overlook this
distinction in analyzing street-level bureaucracies because the line between
formal and informal routines is often very uncertain. For example, the of-
ficial policy of one legal services office may be to accept only emergency
cases. Another office in the same city may have an open intake policy but the
attorneys may informally decide to recruit and work primarily on the most
needy cases. In both offices the policy results are the same, but in one office
they stem from agency intake policy, in the other from the structure of work-
ers' informal assignment of priorities. Often agencies will adopt as official
procedure practices that workers previously adopted informally. When po-
lice departments distribute fire hydrant spray caps for recreation, the new
policy replaces the previous informal practice of overlooking illegal opening
of hydrants by neighborhood residents on hot days.
The following four chapters elaborate some of the routines and simplifica-
tions that arise in street-level work in response to job stresses. Although
these routines and simplifications originate in the coping needs of individual
workers, they nonetheless add up to street-level policy, and they become
the patterns of agency behavior with which clients and policy reformers must
contend.

86
 CHAPTER 7

Rationing Services: Limitation of

Access and Demand

Theoretically there is no limit to the demand for free public goods. Agencies
that provide public goods must and will devise ways to ration them. To ra-
tion goods or services is to establish the level or proportions of their distribu-
tion. This may be done by fixing the amount or level of goods and services in
relation to other goods and services. Or it may be done by allocating a fixed
level or amount of goods and services among different classes of recipients.
In other words, services may be rationed by varying the total amount avail-
able, or by varying the distribution of a fixed amount.
This usage is consistent with the familiar application of rationing in war-
time. During World War II, for example, automobile tires were rationed by
restricting their production for domestic purposes and limiting individual
purchases, making them costly, and establishing priorities among users (doc-
tors were privileged in this respect). This chapter considers rationing in
street-level bureaucracies that has the effect of fixing (usually to reduce or
limit) the level of services. The next chapter takes up rationing that differen-
tiates among clients.
The rationing of the level of services starts when clients present them-
selves to the worker or agency or an encounter is commanded. Like factory
workers confronted with production quotas, street-level bureaucrats attempt
to organize their work to facilitate work tasks or liberate as much time as pos-
sible for their own purposes. This is evident even in those services areas in
which workers have little control over work flow. For example, police often
cannot control work flow because most police assignments are in response to
PATTERNS OF PRACTICE

citizen initiated calls. 1 Dispatchers, however, make every effort to permit

officers to finish one call before beginning another. Officers often take advan-
tage of this practice by postponing reporting the completion of a call until
after they have finished accumulated paperwork. In this way police
officers regularize the work flow despite substantial irregularity in requests
for assistance.
The way in which work comes to the agency significantly affects the ef-
ficiency and pleasantness with which it is accommodated. Official efforts to
influence the flow of work vary greatly. They range from the mild advisory of
the post office providing patrons with information concerning the times
when delays are likely to be longest, to the extreme measures taken by a
New York City welfare office that closed its doors at noon rather than admit a
greater number of Medicaid applicants than could be processed by available
personnel in an eight-hour day. 2
Clearly there are costs to clients in seeking services. In both of the above
examples agencies seek to inform clients of the costs and the problems they
will encounter-in the first instance, if they seek assistance during days
when post office patronage is heavy; in the second, if in ignorance of the situ-
ation they attempt to apply for Medicaid and cannot be accommodated be-
cause of the high intake demand relative to intake workers. In many in-
stances even the failure to inform clients of likely costs in seeking service
constitutes a consumer complaint.
The highest costs are borne by potential clients who are discouraged from
or forbidden access to bureaucratic involvement. While exclusion from
client status is usually accomplished on the basis oflegal grounds, the popu-
lation of the. excluded or discouraged includes many whose exclusion is a
matter of discretionary judgment. The ineligibility of tenants evicted from
public housing, students expelled from school, or welfare claimants deemed
uncooperative depends not on fixed criteria alone, but also on interactions
with street-level bureaucrats.

The Costs of Service

To analyze individual influence it has sometimes proved useful t() recognize

the relationship between citizens' influence and their command of personal
resources such as money, status, information, expertise, and capacity for
work. 3 People who have these resources tend to be more powerful than

88
 Rationing Services: Limitation of Access and Demand
those who do not. When people have them they enhance personal influence.
When workers for public agencies have them they may be used to direct or
subordinate clients or discourage clients from further interactions with the
agency.

MONETARY
Street-level bureaucracies can rarely assign monetary costs for services,
since by definition public services are free. However, monetary costs are
imposed in several instructive instances. In income-providing programs citi-
zens' contributions to the income package may be manipulated as policy.
Medicare patients may be asked to pay a higher deductible before insurance
provisions become operable. Food-stamp recipients may be asked to pay
more for their stamps. The effective taxation of earned income in welfare
reduces the number of people in contact with this street-level bureaucracy.
Clearly differences in monetary costs serve to ration street-level bureaucrats'
services.
Programs sometimes force clients to incur monetary costs that discourage
them from seeking service. Acquiring records from other agencies to es-
tablish eligibility or securing transcripts for appeals can be costly, particu-
larly if travel is involved. Agencies that keep bankers' hours impose mone-
tary costs on working people who cannot appear without losing wages.
Appointments sometimes require parents to seek babysitters. Street-level
bureaucracies that seek to minimize these penalties introduce evening office
hours, or they provide child-care services.

TIME
Just as available time is a resource for people in politics, it is also
a unit of value that may be extracted from clients as a cost of service. Clients
are typically required to wait for services; it is a sign of their dependence and
relative powerlessness that the costs of matching servers with the served are
home almost entirely by clients. It is to maximize the efficiency of workers'
time that queues are generally established. A primary reason that clinic-
based practice is more efficient than home-based practice is simply that it is
patients and not physicians who spend time traveling and waiting. Police-
men also allocate time costs by stopping to question young people who,
while not guilty of any crime, are judged to require reprimanding. 4
Some teachers in some school systems make home visits to meet with
parents, while others schedule parent-teacher conferences after school on
specific days set aside for such purposes. (If there are two parents and one or
both work, both are unlikely to be able to meet with the teacher.) These al-

8g
PATTERNS OF PRACTICE

ternative perspectives on parent-teacher conferences measure significant

differences in the value placed on time of parents and teachers.
Time costs are often assessed by street-level bureaucrats as delay; they are
often experienced by clients as waiting. Bureaucracies can reward clients by
expediting service, punish them by delaying service. Court postponements
can function in this way, as can an increase in the time between intake inter-
views and placement on the welfare rolls. Importantly, bureaucracies often
have little interest in reducing delay, since more expeditious processing
would simply strain available resources.
Assessed time costs may also be experienced as inconvenience, although
they are levied as procedure. For example, when an agency refuses to re-
ceive complaints over the telephone and requires that they be written, it
may cut off complaints lodged frivolously or on impulse, but also discourages
complainants who would protest if it were easier. 5 Requirements to com-
plete multiple forms and produce extensive documentation function simi-
larly. It is possible to make an argument that since the real costs of delay and
elaborate procedures are the activities foregone while waiting, that is, op-
portunity costs, it is justifiable that poor people wait longer than the more af-
fluent, since the opportunities foregone are less valued by the society. 6
However, at the very least this elitist view is based on a calculus to the terms
of which clients have not consented.

INFORMATION
Giving or withholding information is another way in which services
may be rationed. Clients experience the giving or withholding of infor-
mation in two ways. They experience the favoritism of street-level bureau-
crats who provide some clients with privileged information, permitting
them to manipulate the system better than others. And they experience it
as confusing jargon, elaborate procedures, and arcane practices that act as
barriers to understanding how to operate effectively within the system. The
emblematic carrier of this characteristic is the court clerk who runs his words
together in an undecipherable litany to the dominance of court procedures
over citizens' rights. 7 At the bureaucratic level the giving and withholding of
information is most obvious in examining how agencies manipulate their case
loads by distributing or failing to distribute information about services.
Conventionally, analysts assess the demand for services by studying client
rolls and visits. (Demands are statements directed toward public officials
that some kind of action ought to be undertaken.) 8 If it is recognized that
manifestations of client involvement may not fully reflect client interests,

go
 Rationing Services: Limitation of Access and Demand
analysts contrive ways to assess underlying needs, for example, through atti-
tudinal and census surveys. From this assessment administrators and politi-
cians make claims about appropriate levels of services.
However, if it is recognized that organizations normally ration services by
manipulating the nature and quantity of the information made available
about services, then it is easily seen that demand levels are themselves a
function of public policy. Client rolls will be seen as a function of clients' per-
ception of service availability and the costs of seeking services. Client de-
mand will be expressed only to the extent that clients themselves are aware
that they have a social condition that can, should, and will be ministered to
by public agencies.
When New York City reduced acceptance rates for new welfare cases at
seven centers by 17 percent it accomplished this feat by tightening the
application process. This meant not only more careful scrutiny of applicants'
claims, but also more documentation and inquisition was required, which
contributed a separate measure of rationing. 9
This perspective is illustrated by indices of need for legal assistance for
domestic problems. When a sample of Detroit residents were asked if they
required a lawyer for assistance with some domestic-relations matters,
scarcely more than 1 percent answered affirmatively. It would have been dif-
ficult to predict from this survey that approximately 40 percent of the clients
of legal aid and neighborhood law offices originally sought help with domes-
tic problems. 10
Needs become manifest when the institutions that might provide assis-
tance send out signals that they stand ready to assist. The 40 percent of the
clients who originally sought help with domestic matters might have been
only a small portion of the population that could have benefitted from such
assistance. Some who could have used such services may have been deterred
from seeking them. Since legal services are vastly underfunded, even more
dramatic demonstrations of need might have materialized if more lawyers
had been available.
Information about service is an aspect of service. Withholding information
depresses service demands. For example, the campaign to reform welfare by
dramatically increasing the welfare rolls was based on the view that a politi-
cal movement could help overcome the stigma attached by potential recipi-
ents to welfare status. It could provide the information necessary to realize a
substantial increase in the number of recipients. 11 The failure of public wel-
fare agencies to make sure potential recipients receive the benefits to which
they are entitled contrasts dramatically with the success of social security
PATTERNS OF PRACTICE

and Veterans' Administration benefits. The difference is that the clients of

these two income support programs-the elderly, and veterans-are not
socially stigmatized. 12
Client statistics may not indicate much about the objective needs of the
client population but they reflect a great deal about the organizations that
formally cater to those needs. 13 Thus growing demand for adult continuing
education partly exists in the felt needs of the adult population, but the
demand also is responsive to the publicity generated by colleges and univer-
sities and their desire to attract students and their tuition. The demand for
emergency police services exists to an unknown degree, but the introduction
of a 911 central telephone number and dispatch system makes it more likely
that citizens believe the police will respond quickly. After the system is in-
troduced the increase in 911 calls will be responsive to organizational factors
such as publicity about the service and response time as well as more objec-
tive factors such as population growth and changes in the age distribution of
the population.
Although the dominant tendency is for street-level bureaucracies to at-
tempt to limit demand by imposing (mostly nonmonetary) costs for services,
there are some times when they have a stake in increasing their clientele.
They will do this through an analogous rationing process, now directed
toward increasing utilization.
Agencies are likely to try to increase their clientele when they are newly
established and have to prove their ability to put services into operation.
Thus the tripling of service complaints when Boston introduced its Little
City Halls program was particularly welcome by its sponsors. 14 Efforts to
increase clienteles were generally noticeable when central funding sources
launched many subordinate service agencies, which saw themselves compet-
ing for funds in the next fiscal cycle. Such agencies would "beat the bushes"
for clients in order to demonstrate that they were worthy of future support.
Community action agencies and neighborhood mental health centers have
been cases in point.15
Established street-level bureaucracies may also attempt to increase their
clientele if they perceive themselves under attack and calculate that demon-
strations of significant service provision, or increases in clientele, might aid
their cause. Relatedly, street-level bureaucracies may attempt to increase
the number of clients when they are competing against other programs with
similar objectives. Such agencies perceive that they are competing for the
same client pool, and that only the more successful will survive in the next
budget cycle.
This competition also is conducive to quasi-legitimate fraud directed to-

92
 Rationing Services: Limitation of Access and Demand
ward making the agencies look better. For example, when drug treatment
centers were few they could afford to impose rigorous residential require-
ments, particularly since clients' commitment to their own rehabilitation was
considered critical to therapy. When the number of such institutions in-
creased in the early 1970s in response to available funding, and the popula-
tion of drug users started to decline, to increase their clientele the centers
began to relax their enrollment requirements (for example, by accepting
clients who previously would have been judged too difficult to help). They
also relaxed attendance requirements, so that a treatment bed might be oc-
cupied by someone who was not in fact a full-time resident of the center. Be-
sides drug treatment centers, other organizations that have competed for
larger shares of a fixed client pool include mental health centers funded in
the same city, and academic departments competing for students within a
university.
In theory this bureaucratic competition might provide precisely what bu-
reaucracies importantly lack-a substitute for market place accountability.
This, of course, is the idea behind educational vouchers. However, the
healing effects of competition are too often mitigated by the residual bureau-
cratic aspects of the competing organizations. Faculty members in academic
departments with declining enrollments are still protected by the tenure
system, rewards for research (and bringing in research grants), and other fac-
tors that protect them from being assessed solely on criteria of service to
students. Similarly, educational voucher experiments have foundered on
teachers' tenure, union opposition, and parental inability to express prefer-
ences within the system for lack of information on the implications of the
available choices.

PSYCHOLOGICAL
Bureaucratic rationing is also achieved by imposing psychological costs on
clients. Some of these are implicit in the rationing mechanisms already men-
tioned. Waiting to receive services, particularly when clients conclude that
the wait is inordinate and reflects lack of respect, contributes to diminishing
client demands. 16 The administration of public welfare has been notorious
for the psychological burdens clients have to bear. These include the degra-
dation implicit in inquiries into sexual behavior, childbearing preferences,
childrearing practices, friendship patterns, and persistent assumptions of
fraud and dishonesty. 17 Nor have these practices been confined to the "un-
enlightened" 1950s, although some of the more barbaric features of welfare
practice, such as the early dawn raids to catch the elusive "man-in-the-
house," are no longer practiced.

93
PATTERNS OF PRACTICE

To take a modest example, women applying for Aid to Families with

Dependent Children at times are required to submit to an interview with
lawyers, in which they must agree to assist the welfare department in prose-
cuting the father of their children. Apparently many women are unwilling to
agree to this, since it would jeopardize the tenuous but at least partially satis-
factory relationship that they may have with the childrens' father. They fear
that the support they currently do receive and the positive benefits of good
relations with them would be cut off by alienating the fathers, who may not
be making substantial incomes anyway. Applicants are thus forced to lie or
risk the loss of an important relationship. The interviews are conducted in a
legalistic way with little sympathy for the position of the applicant. Many eli-
gible potential clients do not complete the application process, because they
prefer not to suffer these pressures and indignities. 18 Like so many monitor-
ing precedures in welfare, it is unclear if monies recovered through these
procedures equal the costs of engaging in them.
Psychological sanctions serve to reduce the demands from clients within
the system as well as help to limit those who come into it. The defendant in a
lower criminal court who asserts that he or she does not understand the
charges will be silenced by the hostile response of the judge or clerk who
unenthusiastically attempts to redress the complaint. Teachers, by varying
their tone of voice, encourage or discourage pupils from asking questions. A
lawyer in responding to clients can communicate the opinion that the in-
quiry is stupid and the client unworthy of a thoughtful response.
The importance of psychological interactions for rationing service is mani-
fest in the extent .to which clients will sometimes seek or approve of service
simply because they like the way they are treated. Although they later find
against them, sympathetic judges sometimes give thoughtful attention to de-
fendants or complainants with weak cases simply in order to make them feel
that they had their day in court. The reported gratitude of citizens who are
treated in this way may indicate how little people have come to expect from
government. It would seem that clients sometimes judge services positively
if they are treated with respect regardless of the quality of services. In this
connection a study of clients' evaluation of walk-in mental health clinics
revealed that "clinic applicants are satisfied with almost any response [from
staff] at first so long as the emotional atmosphere of the contact is comforta-
ble." 19 While seekers of mental health services may be particularly sensitive
to the quality of initial client-staff interactions there is every reason to think
that these interactions form a substantial part of clients' initial evaluations of
schools, courts, police, and other street-level services where there are no
clearly defined service products to be obtained.

94
 Rationing Services: Limitation of Access and Demand

Queuing

The most modest arrangements for client servicing impose costs on clients.
This is evident in the way clients are arranged, or required to present them-
selves, for bureaucratic processing. Even the most ordinary queuing ar-
rangements--th ose designed to provide service on a first-come, first-served
basis in accordance with universalistic principles of client treatment-im-
pose costs. 20
Queues that depend upon first-come, first-served as their organizing prin-
ciple elicit client cooperation because of their apparent fairness, but they
may ration service by forcing clients to wait. When clients are forced to wait
they are implicitly asked to accept the assumptions of rationing: that the
costs they are bearing are necessary because the resources of the agency are
fixed. They are also controlled by the social pressures exerted by others who
wait. This is one of the functions of the line, waiting room, and other social
structures that make it evident that others share the burden of waiting for
service.
While resource limitations may be unalterable in the very short run, they
are not necessarily immutable. They derive from allocation decisions that
consider it acceptable to impose costs on waiting clients. Costs will not be
imposed upon clients equally. Long lines processed on a first-come, first-
served basis relatively benefit people who can afford to wait, people whose
time is not particularly valuable to them, or people who do not have other
obligations.
Poor people often suffer in such a system. Not only may clients who ap-
pear more affiuent get served first because it is thought that the costs of wait-
ing are higher for them, 21 but agencies often paternalistically develop policy
as if the costs to the poor were nonexistent. A visit to the waiting room of a
welfare office in any inner-city neighborhood is likely to convey the impres-
sion that the Welfare Department assumes recipients have nothing else to
do with their time. Recipients learn the lesson of people who must seek ser-
vice from a single source. Like the telephone company, the welfare depart-
ment is able to pass on to the customer the costs of linking people with ser-
vice. This system also benefits the average client to the disadvantage of
people with extraordinary needs, since initially it has no mechanism for dif-
ferentiating among clients. However, where the injury to people with ex-
traordinary needs is likely to be severe, as in police work or medical
emergencies, the ordering of services is often deliberately structured to
search for and respond to this information.

95
PATTERNS OF PRACTICE

An alternative to the first-come, first-served waiting room or line is the

first-come, first-served queue by appointment. This system is also norma-
tively acceptable and theoretically has the advantage of eliminating many of
the costs of waiting time. In this queue the costs may appear to be reduced
for the average client, but they may still be significant if appointments are
crowded together to insure client overlap, as is typically done in health
clinics and other medical settings. Crowding appointments may be done for
the convenience of bureaucrats whose time is considered more valuable than
that of clients, and who thus are guaranteed a flow of clients even if one
misses an appointment. The costs of such a queue will also be borne by
clients who seek service but cannot afford to wait for it, who are not dis-
ciplined enough to make and keep appointments, or who are not sure
enough of the likely benefits of service to invest in seeking it. What appears
to the street-level bureaucrat as a fair way to allocate time may be seen by
the client in the light of past experiences of bureaucratic neglect and taken as
a sign that the agency is unlikely to be responsive, or that the problem is un-
likely to yield to assistance.
For some clients the costs of waiting may be quite high. In one legal ser-
vices program approximately 40 percent of eligible clients who received an
appointment with a lawyer for the following week did not keep the appoint-
ment. 22 This may have been because the problem dissolved during the in-
tervening time, or because merely talking to the intake worker provided a
degree of comfort. However, it is equally likely that clients who did not keep
their appointments could not keep them but were afraid to say so, were not
organized enough to show up at the appointed time, or faced their legal
problems without professional advice. Or it may have been that the appli-
cants for assistance interpreted the demand to wait for appointments as a
sign that legal services was not likely to be responsive and assumed that, like
other public agencies, it would not in the end prove helpful.
In any event, the day a client appears to seek assistance may be the day
when he or she is most open to help or the street-level bureaucrat is most
likely to be able to intervene successfully. Catherine Kohler Reissman has
written about mental health services in an analogous situation.

It is obvious that the disequilibrium created by a crisis is a powerful therapeutic tool

that is lost if the situation is allowed to degenerate, through po~tponement, into a
chronic, long-term problem. 23

Similar to the queue by appointment is the waiting list; clients are asked to
wait for what is usually an undetermined amount of time until they can be
accommodated. Although it appears to be straightforward on the surface, the

g6
 Rationing Services: Limitation of Access and Demand
waiting-list system has several important latent functions. First, as we have
seen in the case of Boston public housing, a waiting list tends to increase the
discretion of street-level bureaucrats by providing opportunities to call
clients from the waiting list out of tum, or to provide special information that
will permit them to take advantage of ways to be treated with higher prior-
ity.24 Waiting lists also permit agencies to give the appearance of service
(after all, clients are on a waiting list) and to make a case for increased
resources because of the backlog of demand. 25 The waiting list appears to
record the names of potential clients who are seeking service but cannot be
accommodated, although it is obvious to all that many names continue on the
list only because the agency has not attempted to discover who is actively
waiting and who has long since ceased to be interested.
Some social agencies act as if the waiting list usefully filters potential
clients who are truly in need of service and strains out those whose needs are
not substantial and who thus drop off. This system of rationing may also pro-
vide for a period of time in which spontaneous recuperation may occur, again
reserving client spaces only for those who are needy. 26 However, it is uncer-
tain whether continuation on the list is a sign of substantial need or precisely
the opposite, a sign that the potential client is successful enough in managing
the problem that he or sl{e can wait patiently for services.
A queuing arrangement that maximizes the costs to citizens at the expense
of a relatively small number of street-level bureaucrats is employed by lower
courts, which typically require defendants to appear on a given day, but no-
tify them only as to the hour they should appear. In a typical situation fifty to
one hundred defendants, possibly with a friend or member of their family,
must be ready for a hearing or arraignment, with substantial penalties if they
do not appear precisely at the beginning of the session (when their names are
first called). Here they must wait until the judge arrives, and then wait again
while the judge gives priority to defendants in the lockup who may require
attorneys, defendants whose attorneys plead that they have to be elsewhere,
and defendants whose cases require the testimony of waiting police officers,
who themselves are subject to other priorities. Only when these and other
priorities are accommodated will the docket be called in alphabetical or
some other order.
Defendants may be innocent but by virtue of being arrested are judged
guilty enough to pay in time and uncertainty the price that the court exacts
for scheduling cases for the primary convenience of the judge. Although
practices vary from court to court it is typical that defendants will not be told
even approximately when their cases will be called, so that they must wait in
the courtroom, possibly for most of the day, until they receive a hearing. 27

97
PATTERNS OF PRACTICE

The defendant who has waited through such a day has been instructed in the
costs of continued interaction with the court system and must consider
whether exercising rights or even pleading innocent in a minor matter, al-
though legally valid, is worth the time and irritation. Some court systems
have recently recognized that similar problems, including frequent post-
ponements, inhibit witnesses from appearing and testifying in trials. But the
same analysis rarely focuses on defendants and their experiences in court.
This queue by roundup is also typical of jury impaneling, where citizens
are called for a week of service and must sit in a jury room awaiting assign-
ment, often for several days, perhaps never to be called. The system of-
ficially is justified by the fluctuating and relatively unpredictable demand for
jurors, and again is premised on the high value placed on the court's time
relative to citizens' time. To insure that there are always people ready to
serve, more jurors are called than will be required. If the court could toler-
ate a postponement now and then for lack of available jurors, and if jurors
were called to report serially during the week rather than all at once, less
time would be wasted for prospective jurors. But such practices could only
be adopted if the time of prospective jurors were accorded more value rela-
tive to judges' and lawyers' time than is currently the case.
Clients frequently may be quite willing to pay the costs of waiting. Clients
undoubtedly understand that there are times when they will have to wait,
unless bureaucracies hire enough staff to meet peak demand. And since
demand in most street-level bureaucracies is to some degree unpre-
dictable--even schools often have to hire new teachers or shuffie teacher as-
signments after school has started-it would be too costly to provide services
so that waiting would never occur. Waiting becomes injurious and inappro-
priately costly only under certain conditions.
Waiting is inappropriate when it exceeds the time generally expected for a
service. A person may not resent a two-hour wait in an emergency room to
receive a tetanus shot if it is clear that patients with more serious claims are
being served first. But the same amount of time spent waiting in line simply
to hand in forms to renew a driver's license may be exceedingly irritating.
Waiting may also be resented when it involves the violation of an implicit
agreement. Waiting is regarded as inappropriate when clients have made an
appointment, except when the appointment is considered only an approxi-
mation of the time of service (as in the case of office visits to doctors).
Still another situation in which clients resent the costs of waiting arises
when they wait unfairly. Thus if a favored client gains access to service more
easily than others it will be resented by those who are not favored. Some-
times unfairness in waiting time may be so slight as to go unnoticed by

g8
 Rationing Services: Limitation of Access and Demand
clients. A study of black patients in Chicago hospital emergency rooms re-
vealed that compared to whites waiting time was a little more than three
minutes, incurred primarily by claimants with nonemergency conditions
who sought help when the emergency room was relatively busy. But this cost
is not actually trivial. It is worth noting that a modest three minutes or so,for
the 1 , 1 os blacks in the sample alone, would add up to a full working day for
2,619 people on a yearly basis, 28 a measure of one of the costs of institutional
racism for the blacks of Cook County, Illinois.

Routines and Rationing

The existential problem for street-level bureaucrats is that with any single
client they probably could interact flexibly and responsively. But if they did
this with too many clients their capacity to respond flexibly would disappear.
One might think of each client as, in a sense, seeking to be the one or among
the few for whom an exception is made, a favor done, an indiscretion over-
looked, a regulation ignored.
This dilemma of street-level bureaucrats is illustrated well by the legal
services program. Individually, each attorney is obliged by professional
norms to pursue fully the legal recourses available to clients. For impover-
ished clients this presumably means that attorneys should act on clients'
behalf irrespective of cost. Only if this assumption is correct could the provi-
sion of legal services begin to redress the balance of power in the legal sys-
tem, which every observer concedes favors those who command legal re-
sources. But if all clients' legal needs were fully pursued there would be no
time for additional clients. The dilemma is exquisite. To limit lawyers' ad-
vocacy is to deny poor people equal access to the law. To permit unbounded
advocacy is to limit the number of poor people who can have such access.
Only a reconstitution of the legal system could overcome the dilemma within
the current patterns of inequality: either a radical departure in the amount of
subsidies for legal assistance for the poor or a radical simplification of legal
procedures.
When confronted with the dilemma of serving more clients or maintaining
high quality service, most public managers will experience great pressures
to choose in favor of greater numbers at the expense of quality. Their inabil-
ity to measure and demonstrate the value of a service, when combined with
high demand and budgetary concerns, will tend to impose a logic of increas-

99
PATTERNS OF PRACTICE

ing the quantity of services at the expense of the degree of attention workers
can give to individual clients. Street-level bureaucrats, however, may devise
ways to sabotage management efforts to reduce interactions with clients. The
costs of achieving compliance in the face of workers' resistance may some-
times be more than managers want to pay. An example of such worker resis-
tance is related by Robert Perlman in his study of the Roxbury Multi-Service
Center.
Confronted with the complexity and number of demands being made on them, staff
members resorted to shielding themselves fiom the mounting pressures. They ex-
tended interviews to postpone or avoid taking the next client. They scheduled home
visits in order to avoid intake duty. 29
Whether street-level bureaucrats oppose efforts to limit their interaction
with clients, or whether they accept and encourage such efforts as a way of
salvaging an unattractive or deteriorating work situation, is perhaps the criti-
cal question on which the quality ofpublic service ultimately depends. Al-
though street-level bureaucrats may sometimes struggle to maintain their
ability to treat clients individually, the pressures more often operate in the
opposite direction. Street-level practice often reduces the demand for ser-
vices through rationing. The familiar complaints of encountering "red tape,"
"being given the run-around," and "talking to a brick wall" are reminders
that clients recognize the extent to which bureaucratic unresponsiveness pe-
nalizes them.
Routinization rations services in at least two ways. First, set procedures
designed to insure regularity, accountability, and fairness also protect work-
ers from client demands for responsiveness. They insulate workers from hav-
ing to deal with the human dimensions of presenting situations. They do this
partly by creating procedures to which workers defer, happily or unhappily.
Lawyers and judges, for example, generally accept court procedures that
insulate them from erratic client demands. Police officials resist instituting
(or more properly, reinstituting) a beat system because they are apprehen-
sive that officers would become too involved with neighborhood residents,
and thus perhaps engage in biased behavior. For similar reasons they often
oppose assigning officers to the areas in which they reside, and they advocate
reasonably frequent changes in assignment.
Social workers may be unhappy with the requirement to process endless
paperwork rather than spend time providing client services. But whether
happy or unhappy with job routines the fact remains that they serve to limit
client demands on the system. The righteous objections of critics that rou-
tine procedures detract from primary obligations to serve clients are of little

100
 Rationing Services: Limitation of Access and Demand
account, since in an important sense it is not useful for the bureaucracies to
be more responsive and to secure more clients.
Second, routines provide a legitimate excuse for not dealing flexibly, since
fairness in a limited sense demands equal treatment. Unresponsiveness and
inflexibility reinforce common beliefs already present that bureaucracy is
part of the problem rather than the solution, and they further reduce clients'
claims for service or assertions of need.
When routines lead to predictability they may promote a degree of client
confidence. As a public defender lecturing his peers on increasing client
trust advised: "It's better to tell a client you will see him in two weeks and
then show up, than to reassure him by saying, Til stop by tomorrow,' and
never show. "30
But agency practices do not always lead to predictability. When they lead
to delay, confusion, and uncertainty they assign considerable costs to clients.
At times routines established to protect clients are distorted to minimize
contact or services. For example, to insure responsiveness housing inspec-
tors may be required to make more than one effort to contact complainants.
However, inspectors may become adept at telephoning complainants when
they are unlikely to be home or fail to keep appointments punctually. In Bos-
ton this practice "enhanced the prospects of no one being horne when the
inspector arrived-a practice which when repeated thrice, enabled cases to
be dropped. " 31
The significance of practices that subvert predictability, antagonize or ne-
glect clients, or sow confusion and uncertainty is that they are generally
functional for the agency. They limit client demands and the number of
clients in a context where the agency has no dearth of responsibilities and
would not in any way be harmed as an agency if clio:nts became disaffected,
passive, or refused to articulate demands. Any redr ction in client demand is
only absorbed by other clients who come forward, c r by a marginal and insig-
nificant increase in the capacity of street-level bmeaucrats to be responsive
to the clients who continue to press.
It is for this reason that we conclude that stated intentions of street-level
bureaucracies to become more client-oriented, to receive more citizen
input, and t.:> encourage clients to speak out are often questionable, no mat-
ter how sincere the administrators who articulate these fine goals. It is dys-
functional to most street-level bureaucracies to become more responsive.
Increases in client demands at one point will only lead to mechanisms
to ration services further at another point, assuming sources remain
unchanged.

101
PATTE RNS OF PRACT ICE

The logical but absurd extension of the relationshi p between demand and
services is exemplified by the apocrypha l library that reduced costs by clos-
ing down. Yet it is a real problem that increased patronage of libraries, mu-
seums, zoos, and other agencies providing free goods increases their uncom-
pensated costs when they succeed in becoming more attractive.
Undoubted ly there are dimension s of bureaucrat ic practice in which in-
creased responsive ness does not add to workers' tasks. Addressin g clients
politely rather than rudely or indifferent ly is an area in which greater re-
sponsivene ss is not necessarily burdensom e to the work load. Furthermo re,
reorganiza tion may result in increasing the responsive capacity of workers.
However, most increases in responsive ness--doin g more for clients, or even
listening to them more-pla ce additional burdens on street-leve l bureau-
crats, who will subvert such developme nts in the likely absence of any strong
rewards or sanctions for going along with them.
There are times when bureaucra tic rationing is not simply implicit; limit-
ing clientele or reducing services is the agency's stated policy. In response to
reduced budgets or other developme nts that make client-wor ker ratios con-
spicuously high, agencies will reduce the scope of service in several charac-
teristic ways. In reducing services explicitly they will continue to honor the
formal norm of universalistic service patterns.
Street-leve l bureaucrac ies may reduce services geographically. They may
formally narrow the catchment area from which clients are drawn or reduce
the number of neighborh oods served by a program. Alternatively, because
reductions in service are unpopular , street-leve l bureaucrac ies may prefer to
reduce the number of centers, effectively cutting services to some areas
without formally changing anyone's eligibility. When the borough of Man-
hattan, for example, consolidat ed its municipal court system, eliminatin g
district courts in Harlem, it did not formally change access to the court, but
informally it substantial ly increased the costs of using the court system to
Upper Manhattan residents.
Services can be limited in terms of clients' personal characteris tics. For-
mally, agencies can change income eligibility levels. Informally, they may
limit service by failing to print posters in Spanish or by placing notices in
old-age and nursing homes rather than in public housing in order to attract
primarily an elderly population .
Street-leve l bureaucrac ies also can formally or informally ration services
by refusing to take certain kinds of cases. The decriminal ization of drunken-
ness, for example, formally exonerates policemen from dealing with alco-
holics (although public disapproval still places pressure on the police to do
something about drunks). Informally departmen ts can limit the clientele if

102
 Rationing Services: Limitation of Access and Demand
officers choose to ignore public drunkenness, or they can reduce its place in
departmental priorities.
Even when limiting services is not explicitly the function of rationing prac-
tices, service limitation often is not an unintended consequence of bureau-
cratic organization. Street-level bureaucrats and agency managers are often
quite aware of the rationing implications of decisions about shorter office
hours, consolidation of services, more or fewer intake workers, or the avail-
ability of information. Consider, for example, the efforts of the Budget
Bureau of New York City in 1g6g to decrease welfare expenditures. In a doc-
ument remarkable for our purposes the Bureau suggested several ways to
save close to $100 million. 32 In addition to reducing allowance levels, which
would supply the bulk of the savings, the bureau recommended four ad-
ministrative changes. Each would explicitly ration services in some way. A
new intake procedure was proposed that would require applicants to be ac-
tively seeking jobs prior to the intake interview. This would force people to
accept low-wage work, and, it was hoped, "more aggressive utilization of ex-
isting leverage over the employables would ... have a deterrent effect on
applications for welfare. "33 The authors recognized that for this innovation to
be effective a substantially greater capacity of public employment agencies
would be required, but there was no discussion of the costs of achieving this
increase.
More frequent recertifications would be conducted to induce recipients
who were on the rolls but no longer eligible because of changed circum-
stances to initiate case closings. (More than half of all case closings were then
initiated by clients.) This reform would reduce the time between changes in
clients' circumstances and the next reporting period.
Closing seven outreach centers would save some of the costs of running
the centers, but more importantly, "larger savings are anticipated from sec-
ondary effects .... The most important of these is the opportunity to build
up and maintain the maximum legal backlog between intake and eligibility
increasing average backlog from two weeks to a full month." 34 Among other
secondary benefits of center closings, the authors of the recommendations
expected that "the relative inconvenience to the client of self-maintenance on
emergency grants (for which application is normally made at the center more
than once a week) may have some deterrent effect on [those] marginally
eligible for welfare." 35
Finally, stronger management audits would introduce greater uniformity
in the system and provide better checks on welfare employees, who are por-
trayed in the document as more interested in enrolling clients than in con-
trolling welfare costs.

103
PATTERN S OF PRACTIC E

Of equal interest are the strategies considered but not recommended.

These included reducing intake hours, drastically closing intake centers, and
requiring clients to provide increased documentation of birth, wages, rent
payments, and other details of eligibility. While these provisions were re-
jected because they might result in unmanageable backlogs and infringe on
clients' legal right to a response to their application within a month, the
memo clearly recognizes that these measures would deter application rates
by increasing the costs of applying to clients.
Provisions of this memo have been described at some length not because
they are themselves remarkable but because they illustrate awareness at the
agency planning level of the implications of rationing to limit client demand.
It is naive to accept the rhetoric of public officials that their actions have the
incidental effect of limiting or discouraging client demands. Rather, the op-
posite assumption is more useful analytically and more accurate empirically;
namely, that public employees and higher officials are aware of the implica-
tions of actions taken that effectively increase or decrease client demand.
They may deny such intentions publicly, of course, since their jobs require
obeisance to norms of public service. They may not favor such policies per-
sonally, and they may regret that funding limitations preclude being able to
serve more clients. Nonetheless, it is appropriate to assume that public
agencies are responsible for the rationing implications of their actions.
In 1976 New York City introduced administrative controls that were cred-
ited with reducing the acceptance rate for new welfare applicants by half and
terminating 18,000 cases a month. But this was accomplished because eligi-
bles were being turned away "by very negative administration of work and
parent-support rules," and because half of those terminated failed to show
up for recertification, to respond to mailed questionnaires, or to verify school
attendance. Their ineligibility was strictly a matter of difficulty or reluctance
to pay the costs of remaining on the rolls until forced to do so. Meanwhile,
according to an administrator, welfare centers are "overcrowded," "noisy,"
and "dirty." "Some clients wait four to five hours for service and too often are
required to make more than one visit to the center to complete their busi-
ness. In addition, they don't know the names of people who are serving
them." 36 In these and other ways eligible clients are asked to pay the costs of
seeking relie

104
 CHAPTER 8

Rationing Services:
Inequality in Administration

Free public goods and services may be rationed by imposing costs and fixing
their amount. They may also be rationed by allocating them differentially
among classes of claimants. In street-level bureaucracies services are distrib-
uted differentially for at least four interconnected reasons.
First, as mentioned before, to a degree the society wants bureaucracies to
be capable of responding flexibly to unique situations and to be able to treat
people in terms of their individual circumstances. This is particularly the
case for street-level bureaucracies. Teachers are expected to be interested in
the individual child, policemen to he capable of flexible responses, social
workers to be attuned to individual needs.
Second, street-level bureaucrats often want to make an improvement in
their clients' lives. They derive satisfaction from making a difference for
some clients and resist efforts to reduce the discretion that permits them to
have this influence.
Third, and most obviously, bureaucracies are simply often required to dif-
ferentiate among recipients. Everyone is not equally entitled to public ser-
vices. Eligibility, culpability, and suitability for bureaucratic intervention
must all be determined. Indeed, the process of reducing a person to his or
her qualifications for bureaucratic intervention essentially is the process of
becoming a client. 1 In this respect people-processing bureaucracies have
two tasks: to develop an appropriate set of categories in terms of which peo-
ple will be processed, and to map clients in terms of their qualifYing or dis-
qualifYing characteristics.

105
PATTER NS OF PRACT ICE

An appropriate model for differentiatio n among clients in people proces-

sing is "triage." The term has its origins in a battlefield context. It refers to
the decisions of medical personnel to place wounded soldiers in one of three
categories: mortally wounded, with little hope of recovery; lightly wounded,
not requiring immediate attention; and seriously wounded, but likely sal-
vageable if medical attention is provided promptly. By concentratin g on the
last category prudent triaging provides a guide to optimizing the use of med-
ical resources in battle. In current usage triaging refers to any medical con-
text (as in emergency rooms) in which potential patients are assigned differ-
ent treatment priorities. 2
Triaging provides a useful analogy for bureaucratic differentiatio n for sev-
eral reasons. It operates and is mandated in a context of severe resource
shortages where the costs of waiting cannot be passed on to clients except
with unacceptable consequence s. It differentiate s among clients, albeit on a
sound basis. It requires of field personnel considerable discretion that can-
not be reduced to official guidelines. And it clearly operates to the advantage
of some and the great disadvantage of others. This is the case because some
who are triaged as mortally wounded could be saved if they were given at-
tention. It is not that they are all too far gone; it is only that the probability of
saving people in this category is lower than considered optimal for expend-
ing medical resources.
Triage has its counterparts in many street-level bureaucracie s. Building
inspections and police patrols are often concentrated not in the most run-
down neighborhoo ds but in those that are considered still redeemable.
School children are placed in high tracks to provide resources to those who
are considered capable of taking advantage of them. Psychiatrists and psy-
chologists are inclined to accept for treatment people who they think likely
to respond to treatment.
In some situations there are powerful reasons to think that differentiatin g
among clients is in itself harmful. Tracking in schools particularly can be crit-
icized as serving no useful purpose so far as the children are concerned, al-
though it is a powerful tool for making teaching and school managemen t eas-
ier. 3 However, the problem of triaging in public service is not primarily that
it is sometimes destructive. It is that discretionary judgments are subject to
routine abuse.
This introduces the fourth reason differential distribution of services oc-
curs. Aside from whatever overtly discriminato ry practices develop in street-
level bureaucracie s, differentiatin g among clients occurs routinely because
differentiatio n often assists street-level bureaucrats in managing their work
loads, as in the tracking of school children. Or it may help them cope with

106
 Rationing Services: Inequality in Administration
the ambiguities and psychological stresses of their jobs. Client differentia-
tion may take place because, confronted with heavy work loads and ap-
parently impossible tasks, street-level bureaucrats seek ways to maximize
personal or agency resources, or they attempt to succeed with some clients
when they cannot succeed with all. Sanctioned bureaucratic differentiation,
for which triage is the paradigm, is open to the potential that street-level bu-
reaucrats will differentiate among clients for reasons having more to do with
solving or resolving work-related problems than with providing optimal re-
source distribution. In this they commonly introduce precisely the sort of
particularism that modem bureaucracy theoretically overcomes, however
firmly the rules of the agency require universalistic standards of judgment.
I have said that the routines of street-level bureaucracies may well be the
policies of the agency, whether or not they are consistent with agency regu-
lations and standards. To lend substance to this observation some of the pat-
terns of practice by which discretionary judgments in these organizations are
made should be identified.

CREAMING
Confronted with more clients than can readily be accommodated street-
level bureaucrats often choose (or skim off the top) those who seem most
likely to succeed in terms of bureaucratic success criteria. This will happen
despite formal requirements to provide clients with equal chances for ser-
vice, and even in the face of policies designed to favor clients with relatively
poor probabilities of success. Employment counselors, for example, may
send to jobs people who have the greatest chance to gain employment any-
way, to the neglect of people who are more difficult to place. The Upward
Bound program, dedicated to enriching the educational backgrounds of dis-
advantaged high school students, constantly had to guard against projects tak-
ing students whose chances of getting into college were already fairly high. 4
Why does creaming take place, particularly in the face of official opposi-
tion to the practice? In every case of creaming the agency's incentives re-
ward successes with clients, but they provide no substantial rewards for the
risks taken. Since not all potential clients can be served, the reward struc-
ture of the agency is adopted as its implicit agenda in the absence of powertul
incentives to the contrary. If all clients are equally worthy but all cannot be
served, increasing the rate of personal or agency success becomes primary.
The situation is complicated by the fact that the criteria for determining who
is or is not "high risk" or "likely to succeed" are so problematic that there are
few clearcut ways to challenge the worker who may offend selection norms.
From another point of view creaming takes place in circumstances in
PATTER NS OF PRACT ICE

which there are no controls in assessing success. If teachers were assessed by

the rate of progress their students made compared to a predicted rate, then
the high achievemen t students would not necessarily be most highly valued.
But in reality teachers are judged implicitly by the status and accomplish-
ments of their students and thus seek to teach high achievemen t classes or
move to middle-class schools. Similarly, employment counselors would
value the hard-to-place more if they were given more credit for placing
chronically unemployed people than for those temporarily out of work.
These consideratio ns return us to a concern for the ways in which street-
level bureaucrats' performance s are measured.

WORKER BIAS
Differentiation among clients may take place because of workers' prefer-
ences for some clients over others. They may prefer some clients over
others, despite official norms to the centrary, under at least three circum-
stances. Ultimately they all reflect the fact that workers find greater
gratification in interacting with some clients than with others and have op-
portunities to act on these preferences.
First, some clients simply evoke workers' sympathy or hostility. Like the
Israeli customs officials, workers may be inclined to "give the underdog a
break" 5 or may favor clients with similar ethnic backgrounds , as when racial
or ethnic favoritism prevails in discriminato ry decision making. The Boston
Housing Authority workers who tended to favor white elderly applicants
probably were responding to both ethnic and sympathy appeals when they
selectively provided them with critical information.
It would be as much of a mistake to infer that ethnic or racial appeals
always prevail in affecting discretionary judgments as that they never pre-
vail. Bureaucratic norms operate to restrict the range of determinatio ns
made in this way. Thus, black police officers may make particular efforts to
act in role-prescrib ed ways when confronting black citizens. 6 Displaying the
complement ary tendency, white bureaucrats may be more lenient or toler-
ant with black clients out of fear of being accused of racial biases. The report
from San Francisco that black school children tended to receive good grades
and were told that they were doing well in school, but in fact were not learn-
ing at an acceptable rate, is a vicious example of what can happen when
street-level bureaucrats over-react to the potential for biased behavior.
If public officials were simply biased or racist, and if their prejudices were
regularly manifested in behavior, the problem of bias in bureaucracy would
be more pernicious but easier to root out. At the very least it would be easier
to establish policy directives to reduce bias in bureaucracy . But patterns of

108
 Rationing Services: Inequality in Administration
prejudice are more subtle in the modern bureaucracy dedicated officially to
equal treatment. Modern bureaucracy promises to eradicate prejudicial be-
havior through universalistic treatment; when prejudice does occur, it is
more difficult to erase.
A second circumstance of biased behavior is evident when street-level bu-
reaucrats respond to general orientations toward clients' worthiness or un-
worthiness that permeate the society and to whose proliferation they regu-
larly contribute. This is one of the most well-grounded generalizations that
can be made concerning client processing. Juvenile court judges determine
sentencing severity on the basis of the apparent worthiness of the defen-
dant. 7 Policemen make decisions concerning citizens on the basis of whether
or not they display respect. Trauma-team personnel tend to work harder to
save the lives of the young than the old, the high-status citizen rather than
the low. 8 Other emergency room personnel make moral evaluations of
clients and treat them accordingly. 9
Where more than one street-level bureaucrat is involved, as in the cases of
courtroom processing or multi-disciplinary assessments of handicapped stu-
dents with special educational needs, it is often the moral worthiness of sub-
jects that is negotiated in these settings. 10
These observations are consistent with the policies of organizations that
tend to focus on morally favored clients instead of those who most require
their services. Agencies for the blind tend to be oriented toward children
and employable adults although most blind people in the United States are
elderly or near retirement. 11 Voluntary hospitals and private social service
agencies tend to relocate or reduce their community services when a new
group populates the neighborhood for which the services were originally
established.
As suggested above, there is every reason to think that the general evalua-
tions of social worth that inform the society will also inform the decisions of
street-level bureaucrats in the absence of strong incentives to the contrary.
Under what circumstances are these general notions of social worth likely to
play a major role? At least three hypotheses seem plausible. First, when
street-level bureaucrats have to choose among clients, and biased selection
will not incur major costs, general notions of moral worth will prove impor-
tant. Studies of emergency room bias suggest that blacks tend to receive
equal treatment when they have severe injuries, and when the emergency
room is not particularly crowded. But when emergency room personnel are
under pressure, blacks with nonserious complaints are discriminated against
as personnel have recourse to diffuse conceptions of worthiness in decision
making.

109
PATTERNS OF PRACTICE

Second, street-level bureaucrats will fall back on criteria of worthiness

when there is no obvious end point to the degree of intervention. The
trauma team that struggles particularly hard to save the lives of the young
and the more affluent may be applying greater energy because there are no
limits to the degree of dedication they could display. Since theoretically they
should devote maximum effort in all cases, but in practice cannot (for what is
maximum in this instance?), they establish implicit group standards for the
circumstances under which maximum dedication will be displayed. Simi-
larly, teachers are in positions in which there is no limit to the amount of
dedication they could display to their charges. Choosing among students
who are thought to be more worthy of teachers' time is a way of solving the
dilemma of discovering limits to a theoretically unlimited, but practically
limited, dedication.
This is not to deny that for private reasons doctors and nurses may find
saving the life of a child more rewarding than resuscitating an elderly per-
son. Or that teachers may prefer to concentrate on some children rather than
others. However, it is to say that these tendencies, however rooted in the
workers' personal limitations, also answer deeply felt personal needs as-
sociated with the structure of their jobs: to find opportunities for rewards
when success is uncertain and unlimited dedication is not possible because
assignments are open-ended.
Third, street-level bureaucrats will allow diffuse orientations of moral
worthiness to infiltrate decision making when they have explicitly moral
judgments to make. Workers in the legal system, and others with quasi-
judicial functions such as school disciplinarians and parole officers, are
charged with allocating punishments and with the ambiguous tasks of fitting
the sanction to the offender as well as the offense. In such circumstances
considerations of worthiness play a part. Is the juvenile defendant likely to
be deterred from future offenses by lenience or severity? Is the parent fit
to entrust with custody of the children? These questions require moral
evaluations.
The problem is not that moral judgments are made but that the diffuse
moral assumptions of dominant social orientations are likely to in8uence the
decision. Or that dominant values may shape decisions despite competing
normative standards that would provide alternative solutions. These are the
issues for police officers who must determine when loud and boisterous talk-
ing in ethnic neighborhoods or street-gatherings on hot summer nights con-
stitute breaches of the peace requiring intervention. Or consider the police
shorthand that permits officers to draw conclusions concerning the victim's
complicity in sexual assaults: "If he had time to take his shoes off, it wasn't

110
 Rationing Services: Inequality in Administration
rape." 12 Yet we know that coercion is most effective when it is so complete
that the victim does not resist. Or consider the issues confronting middle-
class judges who must decide whether or not to agree to the petition of a wel-
fare department to remove the children of a low-income parent to foster
homes. When a home. for children becomes <'unfit" under law is hardly a
matter for objective determination.
A third circumstance in which street-level bureaucrats regularly display
biased behavior arises when they are able to act on the view that some
clients are more likely to respond to treatment than others. This category of
favoritism is similar to creaming except that the motivation comes not from
the reward structure of the agency (although this may play a part) but from
the gratification that comes with helping people who are thought likely to
respond to help. Psychologists and psychiatrists often favor verbally ori-
ented, middle-class patients because their modes of therapy are most re-
wardingly practiced with such people. Teachers favor children who assimi-
late information easily because they receive more frequent and positive
feedback. 13
Closely related is the possibility that biased behavior occurs because a
group is regarded as unlikely to respond to intervention. Consider the suit of
the National Association for the Advancement of Colored People against the
New York City Board of Education that charged that black and hispanic
problem children were routinely assigned to special schools, yet white chil-
dren with similar problems were helped in their own schools. 14 Here the
NAACP suggests an instance in which the combined preferences of individ-
ual teachers results in institutional racism.

A Comment on the Ubiquity of Bias

The bureaucratic sources of bias discussed here are nurtured in street-

level bureaucracies. Here work is characterized by high degrees of discre-
tion and resource constraints and the need to control clients in order to
process work efficiently. When work is structured as it is in street-level
bureaucracies we come to look fur the need to differentiate among clients, so
much that it seems as useful to assume bias (however modest) and ask why it
sometimes does not occur, than to assume equality of treatment and ask why
it is regularly abridged.
However, since there is an obvious conflict between universalistic norms
and biased practices we need to understand why this contradiction persists

111
PATTERNS OF PRACTICE

in practice. Among the supports of persistent, unsanctioned client differen-

tiation are the following.
1. The context of client differentiation makes performance of all kinds dif-
ficult to assess. In particular it is difficult to assess equity of treatment. In
some instances it is against the law to collect the data that would be neces-
sary to demonstrate patterns of bias. In other instances there are no sound
indicators of service quality, so it is usually impossible to assess workers in
this respect.
z. Norms of equal treatment often function to reduce and minimize ten-
dencies toward bias and otherwise provide powerful myths about the way
services are allocated. Street-level bureaucrats often believe firmly that they
treat all clients alike.
3 When differentiation of clients is observable, it is attributed to formal
or informal policies that are alleged to be in the best interests of the clients
or the best interest of the greatest number. Slow learners are tracked to help
them avoid frustration. Children with physical, mental, or emotional disabil-
ities are segregated in schools to help them avoid the humiliation alleged to
result from exposure to more normal children and to concentrate resources
on their behalf. Alcoholics who are arrested are thought to appreciate the
comforts of a meal and a warm, dry cell. Clients of mental health services are
thought to be able to make good use of these services. Those rejected for
treatment have their time spared since they are thought unable to take ad-
vantage of assistance. There is an ideology of differentiation that at any given
time rationalizes, excuses, and justifies the intervention orientations of
street-level bureaucrats. (For a further discussion of ideological supports for
client differentiation, see chapter 10.)
4 The rule of normality establishes standards of client behavior from
which deviation is measured. This helps to insure that a part of the client
population will be regarded as deviant. Sociologist David Sudnow has ob-
served that public defenders and judges collaborate in sentencing on the
basis of shared expectations of what constitutes a given offense and how it
should be sanctioned. 15 Court personnel attach expectations of moral behav-
ior to particular formal offenses so that if the moral content of an offense is
more or less than expected for the charge, the charge may be changed even
though it accurately describes the conduct in question. In this connection
Maureen Mileski reports that in one lower court in Connecticut, judges sen-
tenced those originally charged with serious misdemeanors as severely as
those who were allowed to plead to lesser offenses and those charged with
higher offenses. It would seem that judges meted saHctions for the behavior
and not the formal charge.1 6

112
 Rationing Services: Inequality in Administration
The rule of normality also helps insure that a part of the client population
will be regarded as requiring or able to benefit from intervention, and a part
will be thought of as unresponsive or unworthy of help. In general, street-
level bureaucrats establish expectations of client behavior, both in terms of
performance and in terms of their interaction with the bureaucracy. Devia-
tions from these standards tend to be differentiated. Legal services lawyers
may be more responsive to particularly cooperative clients. 17 Defendants
who fail to show deference to judicial procedures or the agents of the law
may be singled out for particularly harsh treatment. 1 &. An early set of im-
migration regulations designed to exclude people with mental disorders per-
mitted erratic or "hotheaded" behavior if exhibited by Italians for whom this
was regarded as a cultural characteristic (and therefore normal), but
regarded such behavior as grounds for excluding northern Europeans, for
whom this was designated an abnormal characteristic.
It is probably fair to say that clients will always be differentiated in terms
of their perceived relative normality, regardless of how absolutely receptive
to intervention they are. This provides street-level bureaucrats with the in-
surance that they always perceive a set of clients for whom they are neces-
sary. In a school of exceptionally bright children the teachers would learn
quickly who was or was not more gratifying to teach. In a client population
that was not particularly verbal or middle class, psychologists would still be
able to discover quickly who seemed most amenable to treatment. The client
world as perceived by street-level bureaucrats is probably much like the
children judged to be able to profit from tonsilectomies when this operation
was still popular two generations ago. In an experiment recounted by soci-
ologist Eliot Freidson, 174 children out of 38g were selected by a panel of
physicians as being able to benefit from tonsilectomies. However, when
another group of doctors examined the remaining 215 children, 99 more
were judged in need of the operation. And when still another panel exam-
ined the rest (116), nearly one-half were recommended for the procedure. 19
One wonders whether service professionals would ever regard at least a por-
tion of a client population as totally unable to benefit from their intervention.
Like the existence of routines, there is nothing surprising about street-
level bureaucrats' expectations of normal distribution. The critical question
is, again, how are these expectations related to legitimate objectives? The
massive problem of children excluded from school provides a case in point.
Although their expressed purposes are educational, public schools difleren-
tiate extensively on the basis of behavior, excluding large numbers from the
school population through suspensions and other punishments. 20 Similarly
courts differentiate among people charged with the same offense on the basi~

113
PATTERNS OF PRACTICE

not of their street behavior but of their courtroom behavior. In these cases
conceptions of normality operate effectively to deny or restrict client treat-
ment.
5 Self-fulfilling prophecies contribute to the persistence of bias by provid-
ing spurious confirmation of the validity of differentiation. Greater sur-
veillance of adolescent blacks by the police results in their being arrested at a
greater rate than other portions of the population. This tends to confirm that
black young adults are the primary delinquency problem. 21 Sending people
to prison, where they are exposed to experienced criminals,'where they are
labeled as criminal and are later treated as such by society, helps fulfill the
prophecy that they were the kinds of people who required severe punish-
ment in the first place. 22 Hospital staffs through subtle signals induce pa-
tients to display the behaviors of sick people, confirming that they were
mentally disturbed or physically ill in the beginning. 23 The case of schools is
perhaps best known. Teachers who expect achievement from their pupils in-
teract with them in such a way as to bring out their full achievement poten-
tiaJ.24 Those predicted to do poorly, however subtly, are more likely to fail.
It should be no surprise that self-fulfilling prophecies run throughout
street-level bureaucracies. If clients are differentiated they will respond to
that differentiation by accepting in part the implications of the differentiation
for their own identities. They will also respond to the role of client by con-
forming to bureaucracy's expectations concerning client behavior. From this
interactionist perspective it is hardly surprising that differentiation of clien-
teles and the necessity of involvement with bureaucracy lead to the interper-
sonal dynamics we call self-fulfilling prophecies.
6. In the absence of adequate performance measures and in the context of
making significant judgments affecting clients' well-being, street-level bu-
reaucrats depend heavily on subjective assessments of the validity of their
practices. This tendency is strongly supported by the feeling that the work
they do is so specialized that no one else is in a position to criticize or even
comment on their practices. Police, teachers, welfare workers, and other
street-level bureaucrats consider themselves isolated from ordinary citizens
who cannot appreciate the difficulties and abuse that they experience or the
uncertainty of the rewards. 25
Street-level bureaucrats are receptive to information that seems to con-
firm the legitimacy of their differentiation of the client world and thus sup-
ports their patterns of practice. In this they reftect the general psychological
tendency to receive and incorporate information that is supportive of their
world view and to filter out information that appears contradictory. They also
 Rationing SeiVices: Inequality in Administration
reflect the general tendency to seek information among peers, who may be
expected to be like-minded.
Street-level bureaucrats are conspicuously prone to scan their environ-
ment for empirical validation of their views. Their conceptions of clients
tend to be consistent with perspectives that exonerate them from responsi-
bility for clients' fate. They are particularly inclined to believe that experi-
ence provides the basis for knowledge in assessing the client world. While
validity by illustration is logically indefensible it is a significant social fact
that influences street-level behavior. We may hypothesize that validity by
illustration ("I know it's true because I once had a client who ... ") will
prevail in proportion to the worker's need to cope with the uncertainties of
decision making and the potential consequences of those decisions. The po-
liceman who draws a gun hasti!y because another officer was recently slain
with a knife when he failed to draw his gun has a powerful argument with
which to defend himself when questioned for abridging department policy.
Undoubtedly there are many street-level bureaucrats who refuse to accept
the perspectives of their jobs that arise in the occupational subculture. Still,
the strength of mechanisms adopted to cope with the work is great precisely
because, ifthey are successful coping devices, they work (by definition). The
need to cope acts as a barrier to anomalous information that might challenge
the routines and orientations that have been developed over time. Changes
in procedures are not necessarily resisted because workers are against
change per se, but because change threatens the existence of coping routines
and orientations that serve to rationalize the work. Similarly, anomalous in-
formation is not heard because it contradicts assumptions that make the job
more rewarding or rationalizes its contradictions.
7 Unsanctioned, persistent differentiation is supported by the racism and
prejudices that permeate the society and are grounded in the structure of in-
equality. Differentiation is intrinsic to street-level bureaucracy, but social
inequality supports it and helps account for the cleavages in terms of which
differentiation takes place. Thus the need to routinize, simplify, and dif-
ferentiate in the context of inequality leads to the institutionalization of the
stereotypical tendencies that permeate the society. Whatever prejudices
street-level bureaucrats as individuals do or do not have, the structure of
their work appears to call for differentiation of the client population, and
thus there is structural receptivity to prejudicial attitudes. The need for
simplification exists, so to speak, prior to the stereotype. The stereotype is
nurtured in a context where it functions to divide up the client population. 26
This does not mean that all street-level bureaucrats are prejudiced or that

115
PATTERNS OF PRACTICE

efforts to reduce biased behavior ought not to be promoted. It does mean

that efforts to eliminate prejudiced behavior will tend to yield best results if
they address directly the work problems for which the holding of biases is a
psychological solution. Workshops that help workers discover that the as-
sumptions they hold about clients are not necessary to function effectively,
and those that provide information about techniques of interaction would
likely succeed in eliminating biased behavior far more often than more ab-
stract seminars on race relations.
From this perspective the problem of bias is a profound one, not only for
the quality of service but also for the legitimacy of government. There can be
little official recognition that bias exists if it is bureaucratically functional.
Clients and concerned citizens see biased behavior. Street-level bureaucrats
on a daily basis see attitudes forged from experience reinforced in their va-
lidity. Clients see unfairness; street-level bureaucrats see rational responses
to bureaucratic necessities.
This seems to insure a high degree of conflict over public service delivery.
Citizen groups will continue to spend great energy conceiving ways to bring
accusations of bias to the attention of street-level bureaucracies and design-
ing ways to overcome it. The paraprofessional movement, decentralization,
work force integration, and other reform waves have largely been motivated
by a desire to overcome perceived biases in a context where the issue cannot
be engaged directly, in part for the reasons mentioned here. The problem is
made no simpler by the observation that clients come to expect bias and ne-
glect. That differentiation serves bureaucratic purposes is not acceptable to a
clientele ready to read indiflerence into routine procedures and favoritism
rationalized as service-related.

116
 CHAPTER 9

Controlling Clients and

the Work Situation

Every social order depends on the general consent of its members. Even the
most coercive of institutions, such as prisons, function only so long as those
affected by the institution cooperate in its activities (even if the cooperation
is secured ultimately by force). Typically, cooperation is neither actively
coerced nor freely given, but, rather, it emerges from the structure of alter-
natives.
In the previous chapters I discussed ways in which patterns of street-level
practice function to ration services. A second general function of street-level
practice is not so much to limit services or choose among clients, but to ob-
tain client cooperation with client-processing procedures. The work that
clients are expected to cooperate with may or may not be consistent with
agencies policy declarations. It will, however, be consistent with street-
level bureaucrats' conceptions of how to process work with minimal risk of
disruption to routine practice.
Street-level bureaucrats' need to control clients as well as the incomplete
nature of that control have been discussed earlier (chapters 5 and 6). Here I
consider selected aspects of practice that commonly contribute to routine
control of clients.
1. Street-level bureaucrats interact with clients in settings that symbolize,
reinforce, and limit their relationship. It is practically a cliche to observe that
the severe appointments of a courtroom, dominated by a bench behind
which a black-robed judge looks down at other courtroom participants, con-
vey the power of the system of laws over the individual. Separate entrances

117
PATTERNS OF PRACTICE

for judges, commands to stand whenever the judge arrives or departs, and
the unintelligibility of the court clerk further contribute to the mysteries of
the courtroom.
Each service setting functions somewhat differently, but in their different
ways each contributes to client compliance. Many offices in which people
seek service are structured to separate clearly the workers from the clients
by means of an imposing information desk. 1 Clients, when interviewed, are
led to "offices" that, lacking partitions, violate privacy by permitting every-
one to view (and listen in on) everyone else's work. Fixed rows of desks in
schools, all facing the teacher, physically represent the demand for order
that teachers and schools require. Like uniforms, settings facilitate the func-
tioning of the bureaucracies by drawing attention to the location of power
and cuing the expectations of clients.
These messages are not accidental. They are fostered by the agencies and
generally consented to by the society. It is interesting to observe in modem
courthouses the extent to which the traditional courtroom setting--dark,
polished wood, the bench, separate entrances, the flags, epigrams celebrat-
ing justice-are retained in otherwise nontraditional architecture.
Consider also the tenacity of setting configurations in other public ser-
vices, and the extent to which departures from tradition appear to be radical.
In the public mind a nontraditional school is simply one without fixed desks.
Are clients important and valued as people? Provide them with comfortable
chairs and sofas on which to sit while they wait, ask them if they are comfort-
able, and reassure them if they must wait that they have not been forgotten.
Are clients of little account? Neglect these considerations and have a small,
cramped waiting room with little attention available. It would be mistaken to
think of service settings as accidental. It is often a matter of policy that public
services are able, or consider themselves unable, to plan for client comfort.
2. Clients are isolated from one another. 2 Public service bureaucracies are
organized so that clients have little knowledge of others in the same position.
Most client processing is shielded from the scrutiny of other clients. Isolated
clients are more likely to think of themselves as responsible for their situa-
tions. They are unlikely to see their condition as a reflection of social struc-
ture and their treatment as unacceptable.
When client processing is done in public, the impression is accurately
conveyed that clients are competing with one another for the attention or
favor of street-level bureaucrats. As suggested earlier, in the brutal realities
of triage, clients perceive that they gain special treatment or the attention of
workers only at the expense of other clients. The bureaucratic defense
against special treatment is also germane here: "If I give it to you I would

118
 Controlling Clients and the Work Situation
have to give it to everyone." In street-level settings in which clients do know
each other-in schools, mental hospitals, prisons-<!lient control is fostered
by the competitive systems of rewards, fostering among clients individual
orientations rather than collective solutions to problem solving.
Street-level bureaucracies tend to resist organization by clients when it
occurs. They tend to regard client organizations as unnecessary, frivolous,
likely to be irresponsible, or not representative of clients' true interests.
There are no objective measures of the validity of such assertions. From
some perspectives any or all might be true. However, these assertions are
most usefully regarded as defenses against client organization, intended to
diminish their influence among potential recruits or third parties whose sup-
port is sought, or to lay the groundwork for an intransigent official response.
In the past decade prison inmates, black high-school students, and welfare
recipients all have been regularly subject to such official responses when
they have attempted to organize.
Public officials often prefer to suppress or disorient client organizations
because they can never be sure at what point they will peak or major conces-
sions will be required. However, one lesson learned well by public officials
during the past ten years is that it is often possible and desirable to encour-
age client organizations in order to provide a buffer between individual
clients and the agency. Lacking substantive powers or the resources to act
effectively, client organizations often provide the appearance of access while
actually influencing only those areas in which policy decisions do not materi-
ally affect agency behavior.
3 The services and procedures of street-level bureaucrats are presented
as benign. 3 Actions affecting clients are always taken in their best interest.
Clients are expected to be grateful for benefits they receive. Where street-
level bureaucracies constrain clients who are not regarded as guilty-as in
schools, hospitals, and noncriminal arrests by the police (e.g., apprehension
of alcoholics}-the ideology of benign intervention is particularly necessary
to justify practices of questionable value to both client and worker groups.
When combined with clients' deference to the more extensive education,
training, and expertise of street-level bureaucrats, the ideology that street-
level bureaucrats' intervention is in the interest of clients appears to be a
particularly important instrument of control. 4
4 Clients must come in for service. With a few important exceptions
street-level bureaucracies require clients to appear for service, rather than
have workers go to clients. In part this is a matter of efficiency. As suggested
earlier, the monetary costs of providing service would increase dramatically
if the costs of having to wait were borne by workers rather than clients. How-

119
 PATTERNS OF PRACTICE

ever, there is more to requiring clients to appear on the bureaucracy's turf

than mere economy. Workers face physical and psychological threats when
they leave the safety of the office or service headquarters. As the social gap
between workers and clients widens, as workers come to regard themselves
as more professional, they increasingly resist the home visits that, in many
cases, were the keystones of good practice a generation ago. Doctors and
other medical personnel are now unavailable for home visits. Teachers ask
parents to consult with them at school rather than visit the home. In many
social agencies home visits are on the decline.
When home visits cannot be avoided the society one wa,y or another pays
for asking workers to undertake them. Social workers who must make site
visits to determine the suitability of a home in reports of battered children,
and building inspectors who must visit premises to record violations, often
insist on working in teams to avoid a feared assault in "bad" neighborhoods.
They also arrange to be unable to schedule appointments or find people at
home, and they seek special relief for being willing to undertake hazardous
assignments. 5
The most important exception is the police, whose work, by its nature,
requires interaction with citizens on neutral or hostile turf. The extraordi-
nary routines in which police engage in order to minimize threat in these cir-
cumstances (see below) tend to confirm this view of the importance of the
setting for managing clients.
5 Interactions with clients are ordinarily structured so that street-level
bureaucrats control their content, timing, and pace. Some of this may be
explained by workers' apparent need to collect information and the press of
business that forces them to expedite interactions so as to be able to process
still other clients. Clearly the multiple forms that clients must fill out, as irri-
tating as they are ubiquitous, function to direct clients' energies to the im-
mediate tasks of client processing.
However, the firm direction exercised by street-level bureaucrats over in-
teractions with clients cannot be explained by these factors alone, for client
control is observable in situations in which the next client is not knocking at
the door, and in which the norms of service precisely dictate nondirective
approaches. Carl Hosticka's study oflegal services lawyers' interactions with
clients suggests how much control of clients can characterize client-worker
interactions presumably oriented toward relatively open relations. 6
In general the norms of the legal profession call upon lawyers to advance
and be loyal to clients' interests; to respect and encourage client autonomy;
to insure that clients are involved in and help determine important decisions-
about their cases; and to treat each case individually, so that lawyers respond
120
 Controlling Clients and the Work Situation
to the case at hand rather than to a prior conception of the case. Legal ser-
vices lawyers might be expected to be particularly responsive to these stan-
dards because they work for agencies founded in part to counteract the ne-
glect of poor people by other public agencies and private lawyers, and
because they tend to regard themselves as the legal champions of the
underdog.
However, in actual interactions with clients legal services lawyers tend to
control their clients in ways that undermine these fine standards. In a study
of two legal services offices the following were observed.
Interviews are structured by routines developed to expedite the collection
of information. Lawyers first ask clients questions to complete a general in-
take form, then ask questions designed to complete another, more detailed
form, which the office secretary identified as most likely to be relevant to the
client's case. The dominance of the forms restricts the search process to
those categories anticipated by the format, which must necessarily be gener-
ally applicable to a wide range of clients.
Clients are observed to make repeated efforts to tell their stories in their
own ways, consistent with a nondirective information search process. How-
ever, lawyers continually talk them down by insisting on conducting the in-
terview according to the established format. Lawyers profess interest in
clients' views, but then shut off discussion or questions by asking leading
questions or clarifying small points rather than listening to clients' views.
Data on the structure of the interviews tend to confirm that lawyers domi-
nate. On a measure of dominating the conversation, primarily charting inter-
ruptions of one party by the other, clients interrupted 3.8 times during an
average 35-minute interview. In contrast, lawyers interrupted clients 10.4
times per interview. On a measure focusing on controlling the topic of con-
versation, 94 percent of the utterances oflawyers were directed toward con-
trolling the topic; of these, 8o percent of the utterances were questions, zo
percent of them leading questions. Yet asking leading questions is a poor
way to obtain information in a free and uncontrolled way.
These data reinforce the impressions gained from directly observing law-
yers' interviews with clients. They confirm that interviews are dominated by
routinized procedures in which the prior existence of case types is presumed
and new clients are fitted to the contours of the previously existing types.
Clients are shaped to existing expectations of clients; cases are fitted to exist-
ing views of the kinds of cases typical clients present. All this is justified in
the names of efficiency, thoroughness, and service.
Significantly, the high degree of routinization may dampen the tendency
to differentiate among clients as people. Legal services lawyers do not tend

121
PATTERNS OF PRACTICE

to perceive that clients are significantly different from each other, except in
extreme cases. Only clients who are particularly uncooperative, and those
who are particularly able to assist in managing their own cases, tend to re-
ceive differential responses. Welfare workers also fail to differentiate among
clients. Like legal services lawyers, their interaction with clients is so highly
structured that variations in treatment are not apparent. 7
This is not to say that legal services clients or welfare recipients always
present unique situations. On the whole there is probably a good deal of sim-
ilarity in the situations street-level bureaucrats confront. However, the rou-
tinization of inquiry minimizes the extent to which street-level bureaucrats
can discover unique circumstances requiring flexible responses. Thus we
have the ingredients for another self-fulfilling prophecy. In the expectation
that most clients will fall into previously defined categories, bureaucracies
follow search procedures based on that expectation. Having constricted the
kinds of information they receive, street-level bureaucracies find confirma-
tion that, indeed, clients tend to fall into certain well-defined categories.
6. When control of clients is problematic yet critical to task performance
or personal safety, interactions between citizens and street-level bureaucrats
are dominated by control routines. Some street-level bureaucrats cannot
depend on the setting, or on a set interview process, to control clients.
These workers will develop routines to make client control a precondition of
the interaction.
In teachers' interactions with children and police interactions with sus-
pected offenders, symbols of authority and prior socialization help to insure
client compliance with street-level bureaucracy. However, these social con-
trol mechanisms are often insufficient. Teachers and police officers both
must act immediately to secure the cooperation of the people with whom
they are engaged. Inner-city school teachers, for example, consider main-
taining discipline one of their primary problems. It is a particularly critical
issue in providing educational services when "keeping them in line" and
avoiding phyo;ical confrontations consume a major portion of teachers' time. 8
Even under less threatening circumstances elementary school teachers are
urged to "routinize as much as possible" 9 in order to succeed.
Police rookies are informally socialized by veterans to be tough as a pri-
mary requisite of their training 10 and are regarded as dependable partners in
large measure by their willingness to use force when necessary. Police
officers have most call for mechanisms that will secure compliance with their
authority. They depend on such routines for protection from what might
evolve into a life-and-death conflict without resorting to a show of arms or
other unacceptable deterrents. One way in which they routinize their ap-

122
 Controlling Clients and the Work Situation
proach to interactions is to maximize their safety in the event of an attack.
Thus, for example, police officers physically position themselves to be in the
most advantageous situation if an attack materializes, even if the probability
of an assault is low. 11
Another primary orientation of police routines is to manipulate their pub-
lic image as ready to use violence if necessary. As one observer describes it:
In most threatening situations, the officer attempts to maintain his edge by managing
his appearance such that others will believe he is ready, if not anxious, for action. The
policeman's famous swagger, the loud barking tone of his voice, the unsnapped hol-
ster or the hand clasped to his nightstick are all attitudes assumed to convey this im-
pression. Decisiveness is readily apparent in such a posture, although the officer
himself may have little, if any, idea of what he is about to do. 12

Still another orientation of police routines is to develop capacities for sus-

picion and to conduct patrols to identify people who might be guilty of some
offense or who might pose danger. Police may find clues to the identity of a
potential assailant in a person's walk, dress, style of car, or composure. 13
These routines may be preconditions to effective policing but they also func-
tion to provide an early-warning system for the police officer's control
requirements.
These procedures may work effectively to help street-level bureaucrats
gain control over clients in settings that are not otherwise secure. Certainly
they believe that such procedures are necessary for effective performance
and would be most reluctant to abandon them. Yet there are some good
reasons to think that these control routines help to create a client population
that in unknown ways is different from the one that would exist in the ab-
sence of these protective devices. Again, street-level bureaucrats may be
designing s~lf-fulfilling prophecies in their client-control procedures. By
approaching people in a hostile, abrasive, and suspicious manner they may
evoke the behaviors they predict. The invoking of authority may lead to
rebellion against authority. Thus school personnel may help provoke the be-
havior problems that interfere with their work, particularly if the authority is
not regarded as legitimate.
A person stopped on suspicion may act with hostility toward the police not
because of guilt or intrinsic hostility toward the p<)lice, but out of a sense of
injustice at being stopved. When minority or ethnic communities generally
regard the police as hostile to their interests this sense of injustice might eas-
ily be triggered. The defense of the police (that such approaches are neces-
sary to do their jobs effectively), requires a degree of sympathy for the police
position and a sensitivity to the requirements of the role, which may not be
shared by the client population. Thoughtful police officers apparently make

123
PATTERNS OF PRACTICE

valiant efforts to strike a reasonable balance between an approach that in-

duces compliance and one that accounts for the reactions of those who might
unfairly be stopped. They are aided in many communities by a consider-
able degree of familiarity with police procedures, so that the procedures for
stopping suspects on suspicion are regarded as routine and not taken per-
sonally. Still, the need for control routines represents an ongoing source of
tension between the police and the public.
7 Street-level bureaucrats develop sanctions to punish disrespect to rou-
tines of order. These sanctions are often particularly significant because they
are invoked to affect compliance with bureaucratic order rather than to affect
behavior relevant to service. For example, teachers, like policemen, have
mechanisms that function to provide clues to potential troublemakers and to
exclude from society (in the case of teachers, the society of the school) those
whose offenses threaten the working fabric of the institution. Children who
are suspended for not having a pass, arriving late to class, being absent ex-
cessively, or smoking in the bathrooms are guilty not of educational sins 14
but may find interviews terminated in the rare cases when they are not
willing to conform to the (reasonable) procedures demanded of them by
attorneys. 111 At another point in the legal process judges tend to sanction
defendants on the basis of the seriousness of rule violations, and also on the
basis of their lack of respect for agents of the law. 16
Some insight into the significance for street-level bureaucrats of proce-
dures fostering control over clients can be gained by examining the implica-
tions of threatening to deny workers these procedural coping devices. The
intensity of the resistance of police officers to citizen review boards can be
associated with the fear that people who do not appreciate the pressures and
risks of police work will sit in judgment on officers who do what they have to
do in order to protect themselves. Teachers in traditional schools similarly
fear the removal of the sanctions that, rightly or wrongly, they believe to be
effective deterrents to student misconduct.
At times street-level bureaucrats for some reason cannot call on routine
processing pfocedures. This results in great consternation or exaggerated ef-
forts to appear to be in control. For example, welfare workers in the state of
Washington experience great frustration in processing gypsies because they
have no papers or other means of formal identification to prove birth, death,
marriage, or social security status. Gypsies claim to be illiterate and have the
habit of lapsing into Romany (their own language) when interrogation gets
intense. These welfare recipients are able to evade the normal bureaucratic
processing requirements because they are a relatively small sect whose char-
acteristics are not likely to spread to the general welfare population. How-

124
 Controlling Clients and the Work Situation
ever, they cause great consternation among welfare workers who fear that
their actions might create a precedent for evading the control procedures of
the department. Like parents who can allow exceptions to family rules only
with a warning to their children that negates any implication of generosity or
flexibility, welfare workers typically make such exceptions only with a warn-
ing to their clients that they would be well-advised to learn to read and
write. 17
An analogous process of chastising clients toward whom they are other-
wise being generous is observed among judges who deal out probationary
sentences instead of incarceration. Judges sentencing offenders to probation
tend to chastise and moralize more than when they sentence offenders to
jail. 18 It is as if the judge, guarding against the possibility that the decision
will be called into question by the later behavior of the offender, must com-
pensate for leniency by applying a verbal equivalent of punishment.

Husbanding Resources

Confronted with complex tasks and limited resources, organizations develop

work patterns to conserve the resources available. Managers strive to deploy
resources more effectively or reduce the costs of work processing. They also
may overtly or covertly redefine their objectives, so that what they are trying
to achieve becomes easier to accomplish. These fundamental orientations
toward resource management-efficiency, productivity, and goal clarifica-
tion-are at the heart of most organizational reforms in public service bu-
reaucracies and are discussed at some length below, as they affect the quality
of work.
In street-level bureaucracies, however, practices oriented towards hus-
banding resources arise that are not related to management objectives.
These are the practices developed by workers to conserve their own job
resources. At times they are developed explicitly to conserve resources; at
other times they may be explained by the need to conserve resources but are
not explicitly developed for this reason. Like the other patterns of practice
discussed in these chapters they often have significant implications for the
content of public policy.
Street-level bureaucracies are often subject to unpredictable surges in
demand. This is obviously the case for police departments and emergency

125
PATTERNS OF PRACTICE

rooms. It is also true of welfare offices, schools, and other agencies in which
the work pace would seem more predictable. A client may appear who
presents a particularly complicated processing problem and a few children
may prove particularly disruptive on a given day. When combined with
unexpected but regular breakdowns in agency functioning-such as the
need to cover for a co-worker, or the intrusion of some new regulations
requiring implementation-these unpredictable circumstances require that
street-level bureaucrats have some reserve capacity to respond.
Workers faced with unpredictable surges in work load will attempt to
secure safe time that they can deploy if necessary, time that usually functions
to cushion the work day. "Where workloads or resource supplies fluctuate,
the individual is tempted to stockpile .... " 19 This is the significance of the
police dispatcher's tacit approval of officers who, upon completing an assign-
ment, delay calling in until they have completed leftover paperwork.
Cushioning the work day is rationalized in public services in which there is a
well-recognized need for reserve capacity. Police departments and
emergency rooms may be expected to deploy resources for peak load duty,
but only within certain reasonable limits. In these services there must
always be a planned excess capacity if they are to function effectively. In con-
trast, public services without generally recognized emergency functions are,
if anything, expected to operate above capacity. When schools operate on
double shifts or welfare workers have excessive case loads these institutions
are not considered to be malfunctioning except by a small group of
concerned and attentive citizens.
In such situations street-level bureaucrats have a strong incentive, when
they can, to let the work expand to fill the amount of time available. This
Peter Principle is commonly perceived to result from laziness, inertia, and
inability to plan. While it may be related to these things, it is also rooted in
the bureaucratic need to cope with the unpredictability of demand.
Workers recognize that any sign that they are not working to full capacity
will be greeted with additional assignments, as there are always additional
assignments to be made. Thus there is no incentive to complete an inspec-
tion or an interview quickly, because the time liberated by efficiency will be
reassigned to conduct another inspection or interview. This is not the case
when workers have quotas to meet or business appointments to conduct. In
these cases workers are rewarded by processing assignments more quickly
since they will be able to enjoy or utilize for other purposes the time liber-
ated by their efficiency.
The validity of conceiving street-level bureaucrats as acting to conserve
work time may be judged in part by the vehemence with which workers seek
 Controlling Clients and the Work Situation
to protect their current work loads. Case loads are invariably too heavy. Re-
sponsibilities are always burdensome. These arguments are difficult to in-
spect rationally. Efficiency specialists will seek to disclose an untapped
horde of resources and appropriate them for the agency. Workers' represen-
tatives will demonstrate the good uses street-level bureaucrats make of their
slack time, if they have any-doing paperwork, preparing for assignments,
or taking a breather so as to remain effective-and will demand compensa-
tion if this work must be done while workers are off-duty. For clients slack
time remains a target of apparently greater available resources, of which
both individual clients and groups of clients might take advantage.
The possibilities that workers are overworked and slothful remain. But it
is more helpful to see the problem as structural. Workers have no protection
against the accretion of responsibilities because of incessant demands and
the irrelevance of performance measures of service quality. Given the need
for reserve capacity to meet unpredictable demands, the husbanding of
resources follows.
At the organizational level there are some conspicuously typical ways in
which problems of the need for reserve time are handled. Organizations
relieve their staffs of the need to treat irregular processing problems by de-
veloping special units and special procedures to process problem cases.
These mechanisms are discussed at some length below (pp. 133-139).
Another way in which street-level bureaucrats conserve resources is to
create the conditions for making decisions as free as possible from the im-
plicit pressures of affected parties. For example, when education personnel
in Massachusetts were required to design an educational plan for all children
with special educational needs in the presence and with the active partici-
pation of the parents, these meetings were often preceded by private meet-
ings of the evaluating teams. It is reasonable to infer that the teams met
without the parents in order to present the parents with a united front and to
avoid having to make a decision on the spur of the moment in the presence of
outsiders. 20
Another illustration is provided by Boston housing inspectors. These
workers are required to give complainants a copy of the violations they
record following an inspection. This regulation was introduced to make the
actions of the inspectors visible to tenants who complained and to leave a
record of the visit. However, inspectors are not required to leave with ten-
ants a copy of their "inspector's report," which sets out the actions taken by
the inspector after leaving the premises, including the actions taken on the
violations found. Thus the tenant may receive the impression that the in-
spector's visit was fruitful. In reality the inspector may decide not to take

127
PATTERNS OF PRACTICE

legal action on the violations, and this decision is not taken when the tenant
can review or have access to the process. Moreover, if the inspector decides
that there is no cause for action on the violations, the tenant would only learn
about it if she or he inquired later why no action was forthcoming. 21
Similarly, lower-court judges can be observed to reserve decisions for
later announcement when they think one of the parties to a case would be
highly reactive, or if a worthy but guilty party to a case would be disap-
pointed by a decision.
The desire to make decisions about people in private, particularly when
they are likely to be negative or disappointing, is entirely understandable.
Yet public policy often calls for street-level bureaucrats to make decisions in
public precisely so that they can be exposed to, if not influenced by, the
presence of those affected by the decisions. Such policy incorporates the
theory that clients are likely to become more a part of bureaucrats' reference
groups if they are present at times when decisions are made. Thus it is a mat-
ter of concern that street-level bureaucrats often are able to shift the field of
decision making to a place where clients cannot intrude. The associated ten-
dency to obscure the locus of decision making so that no individual need be
confronted by disappointed or aggrieved clients similarly threatens client
interests.
A final set of practices that function to husband resources have in common
the effective transferring of decision-making responsibility about clients to
other public workers. Street-level bureaucrats do this by permitting lower-
level functionaries to exercise discretion in their place (screening), by ac-
cepting the judgment of others so that independent assessments need not be
made ("rubber-stamping"), and by referrals.

SCREENING
Most people-processing organizations have a formal role for workers who
stand as buffers between street-level bureaucrats and clients. Although
street-level bureaucrats formally make the critical decisions about clients'
status or provide the services, there remains an important formal role for
other workers. ;I'his is to provide information to clients, to determine the
proper slot for clients when discretionary decisions are minimal, or to pro-
tect street-level bureaucrats from inappropriate client pressures (as defined
by the agency).
Thus street-level bureaucracies commonly employ receptionists, clerks,
secretaries, and other facilitators to provide information in person or over
the telephone, to assist clients in locating the proper place in the bureau-
cracy, and to make informal judgments as to whether they should persist in

128
 Controlling Clients and the Work Situation
seeking help. This statement is equally applicable to the receptionist in a
legal services office, a health aid in a VA hospital, an operator in a 911 police
emergency call system, and a secretary of the local public housing authority.
The role of screener would not be cause for comment if screeners per-
formed their jobs as they are defined in theory-that is, making decisions in-
volving minimal discretion. However, in important respects screeners often
come to function as street-level bureaucrats, exercising discretion in impor-
tant areas of people's lives, although without the authority to do so.
Emergency room registration clerks determine the order in which patients
are seen by doctors, and even whether patients will be seen at all. 22 Secre-
taries can be helpful or unresponsive to potential clients, by their manner
conveying whether or not the agency is likely to be receptive, and by help-
ing or not helping potential clients characterize their situations in ways that
will gain them favor. Like registration clerks in emergency rooms, the way
secretaries characterize cases affects the ultimate processing by street-level
bureaucrats. 23
Perhaps an extreme example is provided by the public-housing applica-
tions clerk who identified prospective tenants she considered desirable,
placed them in projects before vacancies became generally available, and in-
formally rejected acceptable applications by placing them in a special file in
the back of the applications file drawer. She did all this while working for an
agency whose formal procedures called for fair and nondiscretionary behav-
ior on her part. 24
Wherever workers encounter the public they are in positions to play the
gatekeeping functions of determining eligibility, conveying information, and
presenting the face of the agency to clients as benign, indifferent, or hostile.
Thus whether the low-level worker's tasks are tightly circumscribed, as in
VA hospitals or police emergency telephone systems, or they present con-
siderable opportunity for the exercise of informal discretion, as in public
housing, these agency buffers can vitally influence citizen access to public
benefits.

RUBBER STAMPING
A different kind of screening is used as the basis for decision making by
street-level bureaucrats who routinely adopt the judgments of others as their
own. Sometimes the views of others determine the actions street-level work-
ers will take; sometimes they will determine the effort workers will devote.
Popularly we call this "rubber stamping," although the process is often more
complicated than this pejorative label implies.
For example, judges commonly accept the decisions of police officers or

129
PATTERNS OF PRACTICE

probation officers in lower-court criminal cases and more or less ratify these
decisions in their determinations. 25 In some domestic relations courts the
recommendations of social workers on the placement of children in con-
tested divorces almost invariably provide the guidelines for judicial action,
although their reports are supposed to be only advisory. The same has been
observed of adoption agency workers and judicial consent to placement. 26
According to Jerome Carlin, in many specialized judicial settings that deal
with the poor "it has become common practice to delegl}te authority for
decision making to administrative personnel: referrees, commissioners, pro-
bation officers, medical examiners, marriage counselors, and others." 27
A particularly significant delegation of authority has taken place in com-
mitments to mental hospitals and in other situations in which mental com-
petence is at issue. Commitments or findings of incompetence often may
take place without hearings or, for that matter, any other proceedings in
which contradictory evidence might conceivably be presented, simply on
the basis of petitions for hospitalization or the reports of physicians. 28 These
practices, however modernized or reformed to some degree in recent years,
provides insight into the abrogation of discretionary decision-making func-
tions under some circumstances.
Judges are not the only ones who accept the recommendations of others in
making their authoritative decisions. Teachers use the judgments of chil-
drens' previous teachers in their informal classroom track assignments, as
early as the second grade, according to one study. 29 Public defenders make
judgments about the worthiness of offenders on the basis of the charges
made by police officers. 30 Emergency room doctors accept a clerk's judg-
ment as to whether an individual requires immediate assistance or is a drunk
and therefore should be treated after other patients, although the "drunk"
may not smell of alcohol and may have serious medical requirements that
would otherwise command the doctors' attention. 31 Perhaps the most me-
chanical application of previously applied labels is that of emergency room
personnel who give immediate attention to all patients who arrive in an of-
ficial vehicle, such as a police car, regardless of their condition. 32
It is fairly easy to understand why bureaucrats would consistently accept
the judgment of others in making determinations. Street-level bureaucrats
confront problems in which they must make significant decisions about peo-
ple and complex situations without being able to interrogate people fully or
investigate the background of their claims. The assertions of other profes-
sionals, who are assumed to know their jobs and are charged with responsi-
bility for making appropriate assessment in their own work, provide signifi-

130
 Controlling Clients and the Work Situation
cant and legitimate cues to decision making in the absence of other sources
of information.
It is entirely rational to depend upon the cues of respected others in mak-
ing decisions. People do this all the time-when asking friends to recom-
mend a mechanic, or a movie, or a mover. Unfortunately, what is rational
private decision making may subvert public policy. Judges, rather than po-
licemen, probation officers, or social workers, are charged with judicial deter-
minations because they are theoretically in a better position to seek and hear
information from all sides, and procedurally in a better position to protect
the participants in a case. From a limited perspective this may be the best
they can do. But when judges pass on responsibility they negate the theoret-
ical safeguards their responsibilities represent in favor of the safety of expert
or informal opinion.
Public policy is also subverted because street-level bureaucrats are gener-
ally obliged to make decisions based upon the case at hand. When clients are
presented with or come accompanied by labels that predict the treatment
they will receive, they do not obtain a response to the case at hand but rather
to the stereotype their label evokes. Thus the "troublemaker" in school, the
"drunk" in the emergency room, and the "rotten apple" in juvenile court re-
ceive responses to their labels and not to the behavior or circumstances that
brought them into association with public agencies in the first place.
Still another difficulty with processing clients through prior identification
by others is that the professionals whose decisions are effectively substituted
for the decisions of those formally charged with authority in these cases are
themselves subject to the decision pressures of street-level bureaucrats. The
probation officers or social workers with many investigations to make and
little chance of making them thoroughly take shortcuts of characterization
and judgment that are similar but not identical to those made by the officials
to whom they report. Thus the social worker is horrified to discover that her
ambivalent, highly tentative report on placing children in a divorce proceed-
ing, rendered under great pressure, highly qualified to reflect the uncer-
tainty of findings, is taken as authoritative and as an appropriate basis for ac-
tion, despite the pleas in her report for further, intensive probing. Hers is
the only information that the judge has to make a decision. The feelings and
actions of both social worker and judge are understandable. But however
understandable, the interests of family members in a fair and humane deci-
sion may be jeopardized. 33

131
PATTERNS OF PRACTICE

REFERRALS
A final set of practices operating to conserve resources is associated with
referrals. Referring a client from one agency to another obviously serves the
client's interest when there is an identified, specific client need and re-
sources are available from the receiving organization. However, there is a
class of referrals which, whatever its contribution to client well-being, ap-
pears to function more to process heavy case loads in resource-poor agencies
than to fulfill specific client needs. Street-level bureaucrats may make refer-
rals as one of the least costly ways to process clients without providing ser-
vices. Thus agencies may maintain benign images of helpfulness and service,
without explicitly having to tum clients away.
This use of referrals is partly a result of the extraordinary demand for
resources relative to the supply. Public agencies, responsibly seeking to
meet clients' needs, attempt to link them with other agencies when their
own resources become swamped. This works to the satisfaction of all when
resources are available in other agencies, but it turns into a referral merry-
go-round when other agencies become similarly inundated. When the Rox-
bury (Boston) Multi-Service Center opened its doors, it expected to provide
black community residents with links to other social agencies to which they
might not otherwise go. However, the agency shortly experienced as many
referrals from other agencies as it was making to other agencies. It may be
inferred that other Boston social agencies, under-staffed relative to the de-
mands on them, saw the new Multi-Service Center as a resource that they
might now exploit. 3 4
Referrals also may represent a way in which agencies protect themselves
by providing symbolic service when actual services are not available. Jeffry
Galper puts it audaciously when he suggests the analogy of the parking
problem in large cities: "there are never enough spots for all in need, but at
any one time some cars are cruising the streets, looking. Referral is a way of
dealing with clients in need without really dealing with them. 35
Referrals also have some of the qualities of court delays and waiting lists.
More people can be accommodated into the service structure at orie time, al-
though no more service is actually provided. And referrals can result in in-
ducing people to stop seeking services because they consider their need less
important now relative to the costs, or they have been encouraged to resolve
their problems on their own. Whatever explains the drop-off from the refer-
ral net, it functions to some extent to ration the community services
available.

132
 Controlling Clients and the Work Situation

Managing the Consequences of Routine Practice

Street-level practices ration service, organize clients' passage through the

bureaucracy, and conserve scarce organizational and personal resources. For
various reasons these practices sometimes prove inadequate, or they evoke
client reactions that cannot be handled through routine procedures. Cases
that deviate from routine processing are not exempt from routinization, how-
ever. Instead street-level bureaucracies call on additional practices to man-
age the first-round costs of processing people in routine ways. These prac-
tices function to absorb dissatisfaction with common procedures, thereby
permitting agencies to continue to process the majority of cases routinely.
Ideally, complex systems ought to have procedures that come into play
when extraordinary circumstances occur. Schools, for example, must have
fire drill routines. However, organizational practices that manage deviance
from routinization often function in ways that, from the perspective of the
client, have little to do with providing optimal services.
Street-level bureaucrats regularly refer difficult or problem cases to other
people employed in their organization. Often this is uncomplicated, as when
novices ask supervisors or more experienced workers to handle clients who
present difficulties. The referral of difficult cases to more experienced work-
ers hardly requires comment. From the point of view of service quality, the
problem arises when referrals are made not because cases defy workers' abil-
ities, but because they interfere with routine procedures. They must be
treated as special by a bureaucracy which cannot afford to hear complaints or
vigorous dissent from decisions at the same time that other clients with simi-
lar claims but less inclination to speak out are also being processed. The
problem is kicked upstairs, not to seek expertise but to manage dissent or
noncompliance. Thus street-level bureaucracies introduce the "pressure
specialist'' 38 to hear and decide on clients who pursue their cases vigorously.
The pressure specialist serves in several ways. Dissenting clients are si-
phoned off, permitting routine procedures to be imposed for the vast major-
ity. Pressure specialists also perform onerous tasks that would otherwise
taint the entire staff. For example, severe punishments in schools are usually
meted out by an administrator or designated disciplinarian, protecting
teachers from having to punish severely students whom they are simulta-
neously asked to instruct.
The availability of a pressure specialist in some respects protects the work-
er from the clientS: strong negative feelings by providing an alternative to
decision making. Rather than listen to clients complain, or worry that a

1 33
PATTERNS OF PRACTICE

decision may evoke client hostility, the worker can process the case through
a pressure referral. Thus the worker's legitimacy is partially protected by the
availability of a channel that places responsibility for difficult decisions in the
hands of others.
Workers can use the availability of pressure specialists to enhance the
prospects of favored clients. For example, welfare workers often take plea-
sure in artfully presenting cases to supervisors in such a way that they are
likely to endorse the worker's judgment. Or street-level bureaucrats can
scuttle clients' prospects without clients' knowledge by giving the appear-
ance of bureaucratic neutrality but privately providing damaging information
to supervisors.
The possibility that decisions can be appealed also enhances the legiti-
macy of the bureaucracy to the client. For this to work on a sustained basis,
however, two conditions must be met. First, and quite obviously, it must
look like channels for appeal are open. Second, and less obviously, these
channels must be costly to use, rarely successful, and, if successful, certainly
not well publicized. The reason for this is simply that if appeals channels
were inexpensive to use or likely to be successful they soon would be used by
clients seeking increased benefits or a favorable disposition. The channels of
appeal would soon be clogged, and the manifest unfairness that some clients
receive more than others because they sought more would undermine the
system. 37
Thus appeals ordinarily require long delays, the services of advocates,
complicated administrative procedures associated with filing, and general
hostility from the challenged agency. 38 Recent innovations responsive to
client pressure often require public agencies to publish the requirements for
appealing and inform clients of their rights to appeal, provide responses
within a specified time period, and offer counsel to clients seeking appeals.
These innovations still require considerable determination and energy from
individual clients.
Public agencies also seek to insure that appeals cannot be sought collec-
tively. The appeals process can function so long as a single client cannot gain
redress for a class of clients. So long as individual clients cannot win benefits
for groups, public agencies can ration the claims oflarge numbers of clients
in many ways, and thus gain protection from an inundation of client
demands.
These observations are generally supported by examining the volume of
appeals in public agencies. For example, through the early 1g6os there was
almost a total lack of appeals from welfare decisions, although federal law
required each state to establish an appeals procedure. In New York City,

134
 Controlling Clients and the Work Situation
where a relatively liberal welfare environment prevailed compared to the
rest of the country, only 15 appeals were taken in 1964, although half a
million people were on welfare at the time. 39
Appeals can also be discouraged by the high probability that they will not
succeed. Allegations of police brutality are rarely made through official chan-
nels because of the conviction that they will not receive a sympathetic hear-
ing from the officers who sit on the hearing boards. In Rochester, for ex-
ample, where 102 complaints alleging "unnecessary force" were registered
in the five- to seven-year period after 1965, only two were upheld by the
police internal inspection office; of the 368 alleging unnecessary force and
other improper behavior, forty-six were sustained. 40
At times street-level bureaucracies institutionalize the pressure specialist,
creating a special unit to deal with cases with which the agency is generally
troubled. In many municipalities landlords are not vigorously prosecuted for
housing violations, in part because judges do not give high priority to hous-
ing cases when cases of apparently greater public urgency-assaults and nar-
cotics, for example-are also brought before them. 41 However, at times the
failure of the courts to prosecute landlords vigorously threatens to under-
mine the legitimacy of the court systems. In such cases local housing courts,
like the one established in Boston, can segregate these troublesome matters
while reserving court time for the regular case load.
Tactical patrol forces provide police departments with special capacity to
allocate officers to high crime neighborhoods and situations with high poten-
tial for violence. Special classes for disruptive students and those with learn-
ing disabilities absolve teachers from a need to deal with these control prob-
lems. This has significant consequences for children assigned to these
classes, as well as for mainstream students deprived of their presence. Even
if a justification for such segregation can be mustered, the highly subjective
nature of classifying students for such classes draws attention to the organiza-
tional rather than the educational functions they serve.
A typical response of many public agencies to the claims generated by mi-
nority and women's rights movements has been to establish special units to
hear citizen complaints and to take responsibility for institutional change in
these areas. Police departments have established internal review boards
(sometimes with outside citizen participation) and community relations units
to present a sympathetic face to the black community. Public school systems
have hired community relations specialists and affirmative action officers to
take responsibility for the complaints of minorities and women and to articu-
late agency perspectives consistent with the interests of these groups. These
steps have contributed to increased minority and female employment in the

1 35
PATTERNS OF PRACTICE

bureaucracies, symbolic rewards to these constituencies, perhaps genuine

changes in the attitudes of some agency personnel, and possibly greater
responsiveness to clients in some circumstances.
However, these innovations also function to protect the bureaucracies
from pressures for change, and they insulate street-level bureaucrats from
the need to confront certain client populations. Police departments channel
what they label minority cases to human relations units when questions
about racial attitudes are raised, freeing ordinary officers from having to
resolve them. The affirmative action office takes responsibility for recruiting
women to the work force, absolving the people who normally do the hiring
from having to change their attitudes about female employees. Moreover,
the ordinary worker recognizes the essentially symbolic and nonintegral na-
ture of the new unit and may display toward it the same antipathy extended
toward the group it is supposed to represent. Thus community relations
officers are correctly made to feel that they are not respected by patrol of-
ficers. 42 Equal opportunity officers responsible for integrating work forces
have to struggle to obtain respect from within the institutions that hire them.
Special units often end up taking responsibility for areas that are properly
the general responsibility of other bureaucrats. They provide a symbolic
approach to deeply devisive issues, and by providing street-level bureau-
crats with a safety valve in their confrontation with clients, they may do as
much harm as good with respect to changing the general orientation of
agency personnel. Sociologist Albert Reiss Jr. puts it simply: " ... the de-
velopment of a special 'human relations' staff, will remove an important
function from the domain of the line worker .... Without specific provision
for implementation in the line, there is little opportunity to apply human
relations to the treatment of clients." 43 Of course police officers can hardly
avoid human relations in the broadest sense. But it is critical that a human
relations unit may absolve them from ultimate responsibility for the way
they behave in certain situations.

EMERGENCIES
The most common structural device for managing the consequences of
routine is the emergency. Emergency practices, whatever their other func-
tions, solve major service dilemmas for street-level bureaucracies.
Most people are generally familiar with some emergency procedures.
Hospitals assign emergency status to people with certain medical conditions
and accordingly take extraordinary actions on their behalf. Ancillary health
protection agencies such as ambulance services and fire departments take
extraordinary actions when an emergency is identified. Police dispatch as-
 Controlling Clients and the Work Situation
signs degrees of priority to calls in which intervention would most likely save
lives or result in the apprehension of felons.
Emergencies are not confined to the reactive, life-saving public services
Americans have become so familiar with through television. New welfare re-
cipients can receive emergency grants upon applying; continuing recipients
can receive them if they are burned out or experience other disasters. Hous-
ing inspections may be expedited if they are classified as emergencies, and in
some cities housing may even be repaired by public agencies if an
emergency is judged to exist. Public housing applicants receive priority if
their need for housing has emergency status. Mental health clinics are in-
structed to take only clients with acute problems who present a danger to
themselves or others. Legal services offices sometimes accept only clients
who have emergency legal needs. It is the rare aid-giving agency that does
not provide emergency assistance, the rare service-providing agency that
does not admit emergency cases to its rolls.
Yet what is an emergency? There would probably be some consensus that
the word denotes a situation that requires prompt attention, threatens the
sustained existence of the subject, and calls for extraordinary actions on the
part of others. 44
However, the use of emergency treatment categories or assignment of
clients to them may not reflect these circumstances. Police emergencies
calling for preemptive responses include situations in which an emergency is
designated primarily because it seems likely that offenders could be caught.
Yet the potential for apprehending criminals is not necessarily related to life
saving. The highest priority is informally assigned to situations in which
officers are the ones whose lives are threatened. Here the highest priority
depends on the occupation of the subjects.
It is relatively easy to find other examples in which emergencies depart
from this definition. As indicated previously, emergency rooms accord spe-
cial treatment to patients who arrive in police vehicles, regardless of their
medical condition. Public housing officers and welfare workers can show fa-
voritism by instructing selected clients how to be categorized as emergen-
cies. New York City developed an emergency response to deteriorated hous-
ing conditions that accorded priority to conditions brought to the attention of
city agencies by rent strikers and other tenant groups. 45 Here emergency
status was affected by who sounded the alarm.
Clearly the category of emergency in public services is organizationally
and situationally determined. A condition is emergent in public services
only if it is called so by an authoritative agent. Patterns usually emerge to
give shape to emergency practice, so that by convention one can often tell

1 37
PATTERNS OF PRACTICE

when a condition is likely to be regarded as an emergency. However, an

emergency does not exist outside of the organizational needs of the agency
and the relationship between the bureaucracy and the clients.
Street-level bureaucracies define emergencies in two ways. They create
the emergency categories, and they determine when clients or cases fit those
categories. In creating the categories they assign priorities but also deter-
mine the reach and limits of their own flexibility. By determining which
cases are emergencies they allocate bureaucratic resources with respect to
formally relevant and nonrelevant attributes of the case.
With the recognition that emergencies are situationally determined, the
function of emergency procedures may be summarized as follows.
1. Emergency procedures aid in the rational allocation of resources. They
are created to direct resources to the most urgent cases. In this respect
emergency measures differ from triage procedures, which seek not the most
urgent cases but the most urgent cases likely to respond to treatment. Most
discussions of emergency procedures focus entirely on considerations of ra-
tionality: how to discriminate among cases most effectively and how to de-
ploy resources most effectively.
z. Emergency procedures facilitate resource mobilization on behalf of
cases. Doctors (and patients if they can) evoke emergency procedures in am-
biguous situations in order to obtain hospital resources. 46
3 Emergency procedures permit treating part of the case load compre-
hensively when resource constraints preclude treating all cases fully. Thus
street-level bureaucracies are able to direct a full range of resources to some
of the case load, thereby preserving a concept of what service could be under
ideal conditions, and salvaging some of the bureaucracy's legitimacy and the
service ideals of agency personnel.
The difficulty with serving part of the case load is that while some receive
special treatment, people with equally severe needs are neglected. Clients
with objectively equal claims do not present themselves to the agencies or
do not evoke emergency procedures, although their condition or circum-
stances may be as severe as those who are treated as emergencies.
4 Emergency categories permit agencies and individual street-level bu-
reaucrats to be selective and to make exceptions. Thus they find one way to
overcome the requirements of fairness and allocate resources according to
private or unofficial conceptions of need.
5 Emergency procedures permit street-level bureaucrats to provide re-
sources to those who threaten the smooth functioning of the agency. For
those who complain or object to the way they are treated, emergency proce-
dures serve as safety valves to deal with contentious, potentially trouble-
 Controlling Clients and the Work Situation
some situations. Emergency status is sometimes accorded clients whose
severe problems would tend to discredit the agency charged with respond-
ing to them. In this case emergency treatment is likely to be forthcoming in
relation to the embarrassment potential of the client's condition or situation.
The client with persistence, higher status, or better access to the media
(through representation by an organized group, for example) is more likely
to have high embarrassment potential than a client who lacks these charac-
teristics.
Emergency treatment contributes substantially to the routinization of
street -level bureaucracies by providing ways to manage some of the negative
consequences of routinization. It provides a routine way to make exceptions
without destroying the legitimacy of the organization. While allocating re-
sources, emergency treatment also permits street-level bureaucrats to exer-
cise a measure of control over their work by selecting some clients for special
treatment. In addition, it protects the agency from criticism and public scru-
tiny by providing a way to deal with cases that present either the greatest ob-
jective or political problems.
For those clients who receive special treatment, of which emergency des-
ignation is often the prerequisite, the public agency is particularly respon-
sive. Yet all clients would like to be treated with the same degree of respon-
siveness. Hence the maneuvering of some clients to be labeled as
emergencies. Individual clients and client-oriented groups have an interest
in expanding the emergency categories so that more and more clients re-
ceive emergency designation. But if all clients are treated as emergencies,
none can receive emergency help. Organizations with responsive
emergency procedures end up designing additional rationing mechanisms to
limit the client demand. Facilities and services become swamped as clients
attempt to receive the special treatment they have momentarily come to ex-
pect. At times the disorder that ensues can result in an enlargement of
public service responsibilities. Emergency rooms take on additional person-
nel and welfare offices become less secretive about emergency grant proce-
dures when clients become educated about their rights.
However, if clients remain nonvoluntary and dependent with respect to
street-level bureaucracies, the limits of this creative disorder are likely to be
reached shortly. Organizations with responsive emergency procedures end
up designing additional rationing mechanisms to limit client demand, or
they restructure emergency services in order to absorb the demand without
threatening organizational functioning. 47

139
 CHAPTER 10

The Client-Processing
Mentality

The drill sergeant who insists that soldiers stand tall, keep their eyes
straight, and march in precision achieves results without knowing the state
of mind, predispositions, or previous military experience of the recruits. He
is untroubled by the needs of individuals and is at ease with mass processing.
Street-level bureaucrats are not so favored. Their work involves the built-in
contradiction that, while expected to exercise discretion in response to indi-
viduals and individual cases, in practice they must process people in terms of
routines, stereotypes, and other mechanisms that facilitate work tasks.
Workers defend these patterns psychologically. They regard their adapta-
tions to the job not only as mechanisms to cope with resource limitations,
but also as functional requirements of doing the job in the first place. Thus
what to critics seem to be compromise solutions to resource constraints may,
from the workers' perspectives, be desirable and necessary components of
the work environment. To attack the routine is to appear to attack the struc-
ture. Clients who challenge bureaucratic routines are taught this lesson
when administrators act to control them or respond defensively to questions
about agency procedures.
However, this does not entirely explain how workers cope or exhaust the
types of psychological adaptations apparently required by these jobs. For
one thing, it does not explain how street-level bureaucrats rationalize the
discrepancy between service ideals and service provision. At least two addi-
tional perspectives on the psychology of street-level work must be consid-
ered in accounting for street-level bureaucrats' persistence and relative job
satisfaction.
 The Client-Processing Mentality
First, street-level bureaucrats modify their objectives to match better
their ability to perform. Second, they mentally discount their clientele so as
to reduce the tension resulting from their inability to deal with citizens ac-
cording to ideal service models. In short, street-level bureaucrats develop
conceptions of their jobs, and of clients, that reduce the strain between
capabilities and goals, thereby making their jobs psychologically easier to
manage. 1
This is particularly significant because street-level bureaucrats' views of
their work, and of clients, are matters of great public concern. Street-level
bureaucrats are often accused of being biased against particular racial or eth-
nic groups or they are thought to be particularly cynical or unreliable in ful-
filling obligations toward particular social groups. The proposal that workers'
attitudes in large part are formed in response to their work setting contra-
dicts some popular views. Popular wisdom often identifies the source of
workers' attitudes toward clients and their jobs in prejudices acquired in up-
bringing and social background. Such perspectives lead to recommendations
to hire better educated personnel or provide further education and training
in public and human relations.
All too often such perspectives fail to take account of the influence of
street-level bureaucrats' work on their attitudes. It is apparent that street-
level bureaucrats change their attitudes from the time they are recruited to
the time when they begin to experience work problems. Differences in the
class backgrounds of recruits tend to disappear in training and trainee sociali-
zation. 2 Furthermore, there is evidence that educational background, which
is closely related to class, is not an important predictor of the attitudes of
workers who experience extreme job stresses. In this connection, sociologist
Eliot Freidson has reviewed studies relating doctors' educational back-
ground to performance and concludes: "There is some very persuasive evi-
dence that 'socialization' does not explain some important elements of pro-
fessional performance half so well as does the organization of the immediate
work environment. 3
This is not to say that biases toward clients do not intrude in street-level
work. However, focusing on the social backgrounds or experiences of work-
ers will not yield a persuasive theory of bias in street-level bureaucracy.
Such a theory should account for the development and persistence of atti-
tudes as well as their direction.
Taking a different view, the origins of bias in street-level bureaucracies
may be sought in the structure of work that requires coping responses to job
stress. Attitudinal developments that redefine the nature of the job, or the
nature of the clientele to be served, function in this way. Considering the
PATTER~S OF PRACTICE

structure of work helps explain the persistence of biases and the difficulties
inherent in interrupting them.
However, the content of coping responses may well reflect the prevailing
biases of the society. The need for biases may be rooted in the work struc-
ture, but the expression of this need may take different forms. Stereotyping
thus may be thought of as a form of simplification. While simplifications are
mental shortcuts (of many different kinds) that summarize and come to stand
for more complex phenomena, stereotypes are simplifications in whose va-
lidity people strongly believe, and yet they are prejudicial and inaccurate
as summary characteristics for groups of people with nominally similar
attributes.
This approach to analyzing the client-processing mentality detaches the
existence of attitudes toward clients and jobs from the content of those atti-
tudes. It suggests that attitudinal dispositions will be rigid or flexible in large
measure according to the degree they help workers cope with job stresses.
On the other hand, it suggests that workers' attitudes and resulting behavior
may be challenged and helped to change if: incentives and sanctions within
the structure of the job encourage change; the structure of the job is altered
to reduce workers' needs for psychological coping mechanisms; it can be
shown that workers can cope successfully with job stresses without depend-
ing upon undesirable simplifications; efforts are made to make simplifica-
tions conform to actual job requirements rather than to unrelated biases.
These general guidelines are grounded in recognition that the persistence of
inappropriate attitudes is related to the work experience, and they can best
be helped to change by focusing attention on the requirements of work.
The following sections treat in greater detail the tendency of street-level
bureaucrats to cope with job stresses by modifying their conceptions of work
and their conceptions of the clientele to be served. At the same time they
show the relationship between attitudinal coping responses and the patterns
of practice that the attitudes support.

Modifications of Conceptions of Work

TENSIONS BETWEEN CAPABILITIES AND OBJECTIVES

Withdrawal from work is one way that people respond to job stress. They
may withdraw in fact, or they may withdraw psychologically. At the ex-
treme, the tension between capabilities and objectives may be resolved by
 The Client-Processing Mentality
quitting. Or, in anticipation of this tension, people may decline to apply for
public employment in the first place. Idealistic young teachers quit because
they cannot tolerate the pettiness of their supervisors or their inability to
teach as they would like or were trained to teach. Zealous young attorneys
leave jobs as public lawyers in despair over making an improvement in the
lives of their poor clients. In some ways these idealists are potentially the
most dedicated public employees. In other respects they are least suited to
do the work. In any event public agencies are left with a work force least
bothered by the discrepancies between what they are supposed to do and
what they actually do.
They and others who withdraw from the work force mute the extent to
which withdrawal behaviors are evident in street-level bureaucracies. Thus,
adaptive attitudes developed may be more moderate than would be the case
if those least able to cope had remained on the job.
Those who do not actually withdraw from the work force may withdraw
psychologically without actually quitting, rejecting personal responsibility
for agency performance. The outward manifestation of these withdrawal ori-
entations are familiar to managers and people attentive to labor-management
relations: absenteeism, high turnover, goldbricking, slowdowns, and general
withdrawal from involvement. These reactions are all outward signs of atti-
tudinal responses to the sometimes overwhelming and insuperable difficul-
ties of gaining gratification in task processes- and achievement. At base-are
psychological developments that function to help workers maintain a dis-
tance from their failure or inability to realize the symbiotic goals of personal
gratification and task realization. 4
The problems of actual or psychological withdrawal from work are compli-
cated in street-level bureaucracies by several considerations. There are nu-
merous incentives outside the job context itself that operate to reduce ~he
extent to which workers leave public service. Civil service systems protect
against arbitrary management decisions, but they also increase the costs of
firing workers or taking actions against them. In addition, workers accrue
rights by virtue of their tenure in public employment, providing powerful
incentives to remain in jobs despite low or declining job satisfaction. For ex-
ample, the right to retire after twenty years' service, or pension rights that
increase with tenure, encourage street-level bureaucrats to remain in jobs
despite the inherent pressures. 5
Indeed, it is possible to argue that these and other conditions of public
employment, when combined with the difficulty of measuring job perfor-
mance, are powerful enough to reduce workers' contributions to agency
objectives to an absolute minimum once a degree of seniority has been
1 43
 PATTERNS OF PRACTICE

achieved. The cynical view is that public workers have very little incentive
to perform. However, while some street-level bureaucrats may retire on the
job, the vast majority continue to be reasonably dedicated to occupational
objectives as they come to define them. 6
In addition to the usual material and psychological incentives operating on
the job, street-level bureaucrats often enter public service with some inter-
est in client-oriented work, embrace professional orientations that call for al-
truistic behavior toward clients, and continually interact with clients, thus
regularly confronting client characteristics and concerns. Mqreover, street-
level bureaucrats do not abandon agency objectives entirely because the
discretionary nature of their jobs and the organizational milieu in which they
work encourage them to develop private conceptions of the agency's objec-
tives. They strive to realize these moditied objectives and measure their day-
to-day achievements in terms of them. They rationalize ambiguities and con-
tradictions in objectives by developing their own conceptions of the public
service (which they may share with other workers). Taking limitations in the
work as a fixed reality rather than a problem with which to grapple, street-
level bureaucrats forge a way to obtain job satisfaction and consistency be-
tween aspirations and perceived capability.
Accepting limitations as fixed rather than as problematic is significant for
two reasons. First, it discourages innovation and encourages mediocrity. It is
one thing to say that resources are limited, another to say that the practices
arising from trying to cope with limited resources are optimal. Yet the ten-
dency to equate what exists with what is best is strong when patterns of prac-
tice must be defended psychologically to avoid confrontations with work fail-
ures.
Second, as I have argued, organizational patterns of practice in street-
level bureaucracies are the policies of the organization. Thus, workers' pri-
vate redefinition of agency ends result directly in accepting the means as
ends. Means may become ends in other organizations, but lower-level work-
ers rarely have as much influence on the drift in goals as in street-level bu-
reaucracies.

PRIVATE GOAL DEFINITIONS

As we have seen, individual workers develop procedures to allocate re-
sources efficiently. Some of these practices are approved or indulged by
their organizations, others are unsanctioned. Parallel developments occur in
conceptions of the work to be done. Just as organizations confronted with dif-
ficulties in achieving objectives may retreat on objectives in order to obtain a

144
 The Client-Processing Mentality
better fit between their capabilities and goals, 7 so too workers can and do
modify their conceptions of the job in order to close the psychological gap
between capabilities and objectives. Thus judges may be oriented toward
punishment and deterrence or corrections and rehabilitation. Teachers may
be oriented toward classroom control or toward cognitive and personality
development. Police officers drift toward concerns with order maintenance
or law enforcement. 8 Possessing a simpler concept of the job than the one
theoretically prevailing in reality, street-level bureaucrats are able to fashion
an apparently more consistent approach to their work.
Street-level bureaucrats also impose personal conceptions of their jobs
when they make superior efforts for some clients, conceding that they cannot
extend themselves for all. At times this perspective results in favoritism
toward certain social groups, but it may also apply without group bias. A case
in point is the public defender who must select only a few cases to push to
trial, settling the others as best he or she can. 9 Teachers similarly rationalize
their inability to pay close attention to all children by drawing special satis-
faction from the progress of children who do receive particular notice.
In these cases efficiency is still the norm and effective triage is again the
ideal. But the benefits gained from modifying goals to make them consistent
with serving a few, when not all can be served well, are not public benefits.
On the contrary they are enjoyed mostly by the workers (and presumably by
the clients who receive special attention). Moreover, they are not open to
popular judgment or normally available for policy analysis. The individual
street-level bureaucrat is not, in a sense, free to abandon private conceptions
of the job without taking on still more of the tensions that go with it. Because
these personal conceptions are adaptive responses they tend to be held
rigidly and are not open for discussion.
The patterns of practice developed by individual workers often only make
sense in the private conception of the job held by the worker, while super-
visors and the public still expect allegiance to a more complex set of goals.
For example, a police officer who fails to make an arrest upon observing an
unlawful incident may strike an observer as negligent. But if the officer pri-
vately understands his or her job to be one of maintaining order and commu-
nity harmony, with law enforcement in the neighborhood a secondary mat-
ter, this behavior may be acceptable according to the officer's private
definition.
In the same way, a teacher who spends a great deal of time with a few
students will not consider fair any criticism of this practice if he or she
defines the job as, at best, the provision of sufficient attention to a select

145
PATTERNS OF PRACTICE

group. It is difficult to investigate conceptions of the job and trace their rela-
tionship to performance. Yet this may be necessary if one would try to
reorient street-level bureaucrats in their work.
Private conceptions of the job have their counterparts in official policy. In
some cases agencies themselves solve workers' problems by imposing a par-
ticular orientation on the work. At other times, the adaptive defensive atti-
tudes of street-level bureaucrats toward their jobs are incorporated in the
service orientation of their agencies although still officially unsanctioned.
Thus the staff of some schools develop collective perspectives on their work
and some police departments develop a shared view of patrol practices, con-
trary to the preferences of supervisors. Recruitment of like-minded people
to the service contributes to collective adaptation to bureaucratic stresses by
excluding staff members who would challenge work-force goal consensus. 10

SPECIALIZATION
Specialization of function in bureaucracy is usually treated as fostering ef-
ficiency, permitting workers to develop skills and expertise and concentrate
attention on their work. For some analysts specialization is synonymous with
modem bureaucracy . 11 Specialization is frequently and increasingly char-
acteristic of street-level bureaucrac.~s. Welfare departments separate social
services from eligibility determinations. Legal services agencies separate in-
dividual client servicing from law reform units. Schools breed educational
specialties.
Like other contributors to efficiency, specialization solves problems for
workers as well as for their organizations. In particular, specialization per-
mits street-level bureaucrats to reduce the strain that would otherwise com-
plicate their work situation. A lawyer in a law reform unit need not balance
the demands of incessant case-load pressures, while his or her colleague who
has high case-load assignments is relieved from considering the larger issues
that clients' cases present. The social worker concerned with eligibility is re-
lieved of concerns for clients' social integration, while the income mainte-
nance worker need not worry whether clients receive undeserved support.
It is undoubtedly appropriate for some workers to be trained in areas that
others are not trained in. Not every teacher, for example, need know French
or Hebrew or Chinese for schools to provide training in languages other than
English. But some specialization relieves other workers from developing
skills they should have. As I have suggested, community relations specialists
relieve others of responsibility for concern with treatment of minorities.
Special community advocates may function to relieve others of responsibility
for being advocates themselves. Even the case of language specialization is
 The Client-Processing Mentality
not so obvious as it might first appear. For should not all teachers in some
city schools know Spanish to be able to converse with a large proportion of
their students? Why should the Spanish teachers and the teachers of His-
panic background have responsibility for communicating with Spanish-
speaking students? Specialization in this case relieves the other teachers of
an important complication in their work lives.
Specialization permits street-level bureaucrats to avoid seeing their work
as a whole. Once specialized they are expected, and expect themselves, to
pursue an agenda that calls for the deployment of a restricted set of (perhaps
highly developed) skills toward the achievement of a result defined by those
skills. Specialists tend to perceive the client and his or her problems in terms
of the methodologies and previously established processing categories that
their training dictates. 12 Rare is the specialist who retains a comprehensive
conception of the client and the alternatives available for processing. In
some fields, such as special education, critics have advocated the training of
general specialists capable of working with children with any learning dis-
ability or physical or psychological behavior. (This confirms the obvious:
teachers should be well trained for the job, and the base of practice and
theory from which they should operate has expanded significantly.)
Public institutions generally have conflicting or ambiguous goals for good
reason. They embrace ambiguity, contradictions, and complexity because
the society is unable and unwilling to abandon certain fundamental aspira-
tions and expectations in providing public services. Specialists undoubtedly
bring important skills and orientations to organizations that cannot develop
them in their staff as a whole. Yet specialization and task specificity should
be analyzed to discover those circumstances in which the costs of relieving
street-level bureaucrats from contradictions and ambiguities may be higher
than the benefits.

IDEOLOGY AND MILIEU

Another dimension of goal consolidation is provided by the occupational
or professional ideology that governs street-level bureaucracies. Ideology
provides a framework in terms of which disparate bits of information are
stored, comprehended, and retrieved. 13 In street-level bureaucracies
ideology also can serve as a way of disciplining goal orientations when many
goals compete. When a school becomes an open classroom school or re-
verts to a traditional model the directors are saying something about their
goals as well as their methods. The same is true in the case of correctional
facilities that assert the primacy of custody over treatment. 14 By stressing
some objectives over others, administrators partially solve the problem of

147
PATTERNS OF PRACTICE

what kind of institution they will run. Thus hiring becomes more rational
because objectives are clearer, and employees have a clearer sense of what
they are expected to achieve.
In recent years considerable attention has been devoted to the trend
towards "medicalization" of social problems. Advanced by physicians and
supported by a public anxious to think that there are "solutions" to behav-
ioral "problems," the medical model has intruded into the worlds of educa-
tion and corrections, and other environments in which human development
is at issue. This trend has been correctly understood as undermining the po-
litical and social status of individuals, who, labeled "diseased" or "sick," are
expected by the society to accept others' definitions of their circumstances
and means for recovery. The significance for social control is substantial.
What in other times might be understood as rebellious behavior may now be
processed as mere sickness, implying no indictment and certainly no cul-
pability on the part of social institutions that may have contributed to the
genesis of the behavior.
Why has the medical orientation become so prominent? The influence of
physicians and the high regard in which most people hold them surely pro-
vides part of the answer. But this does not fully explain the attraction of the
medical orientation to say, educators, who in some respects have competing
professional perspectives.
A substantial addition to understanding the attraction of the medical mi-
lieu in education, corrections, and other fields may be gained by recognizing
the ways in which the introduction of a therapeutic milieu contributes to
simplifying the goal orientations of public service workers. It provides a
defense against personal responsibility of the worker by resting responsi-
bility for clients in their physical or psychological development. It provides a
theory of client behavior to help explain the complex world of the street-
level bureaucrat. And it provides a clear statement of clients' problems in
terms of which responses can be formulated. The hegemony of the medical
model may be explained not only by the influence of physicians but also by
the way it helps street-level bureaucrats solve problems of goal complexity.
This is not to say that goal clarification and reconstruction of work objec-
tives have no value. Schools that assert that reading is primary may be able
to achieve results that elude schools with more diffuse goals. There are un-
doubtedly physiological dimensions to deviant behavior in some instances,
although the pharmacological cure is sometimes worse than the disease. The
question is whether or not public institutions make their objectives and ori-
entations manifest and the costs of their choices clear, and whether or not it
is appropriate to abandon some goals or concentrate more on others.
 The Client-Processing Mentality
DEFENSES AGAINST DISCRETION
Street-level bureaucrats sometimes cope with their jobs by privately mod-
ifying the scope of their authority. Imposing restrictions on the scope of
their powers frees street-level bureaucrats from perceived responsibility for
outcomes and reduces the strain between resources and obj ...ctives.
Denying discretion is a common way !o limit responsibility. Workers seek
to deny that they have _influence, are free to make decisions, or offer service
alternatives. Strict adherence to rules, and refusals to make exceptions when
exceptions might be made, provide workers with defenses against the possi-
bility that they might be able to act more as clients would wish. "That's the
way things are," "It's the law," and similar rationalizations not only protect
workers from client pressures, but also protect them from confronting their
own shortcomings as participants in public service work. 15 At times these as-
sertions are best understood as strategies to deflect clients' claims. But at
other times they are best understood as rigidly held attitudes that partially
have their origins in, and are bolstered by, distress over the gap between ex-
pectations and perceived capability.
Agencies often impose rigidities on their workers. For example, in the late
1g6os, when the welfare rights movement began to pressure welfare workers
to make discretionary grants to large numbers of recipients, welfare depart-
ments throughout the country eliminated discretionary special-grant awards
for furniture and other items. Thus the departments removed from workers
a discretionary option. This circumscribed their power but also eliminated
the tension between the workers' desires to help clients and their need to
control disbursements.
Another way in which agencies help solve employees' role tensions is by
extensively promulgating rules specifying official procedures. From the
point of view of reducing role tensions it is less important that rules are not
necessarily followed than that they are available as authoritative materials
with which street-level bureaucrats can renovate job conceptions to better fit
work realities. Thus rules not only order work but also function to order
workers' role conceptions. 16
Earlier chapters have focused attention on street-level bureaucrats' devel-
opment of work routines to process clients and otherwise treat their respon-
sibilities. These routines often represent more than mere instruments of ef-
ficiency. Street-level bureaucrats also develop attachments to modes of
practice. They appear to feel that their jobs require the routines. In some
street-level bureaucracies,. routines of practice become so dominant that
workers seek to negotiate the routines rather than to obtain the objective for
which routines were presumably developed.

149
PATTERNS OF PRACTICE

Legal services lawyers, for example, have been observed to discourage

clients from raising questions and penalizing those who refuse to follow the
preferred procedures. Similarly, welfare workers have been observed to dis-
favor clients who do not permit them to conduct interviews according to
standard formats. 17 These and other examples of rigid adherence to pro-
cedure suggest the significance for workers of pursuing means instead of
ends.

DEFENSES AGAINST BUREAUCRACY

Earlier chapters have also stressed the tenacity of street-level bureaucrats
in resisting efforts to limit their discretion. They may assert discretionary
dimensions of their job to a greater degree than called for in theory in order
to salvage a semblance of proper client treatment as they define it. Typically,
they develop conceptions of their job that focus on good treatment of some
rather than inadequate treatment of all.
Most of the time escapes from bureaucracy tend to favor some kinds of
clients over others. (This tendency is discussed later in the treatment of
modifications of conceptions of clients.) Sometimes the escape from bureau-
cracy appears simply as a refusal to accept the decision-making formulas of
lite work. The social workers who started making home visits rather than
doing intake, because they telt that additional clients could not be well
served by the agency, illustrate this inclination. 18
Another dimension of the escape from bureaucracy is suggested by street-
level bureaucrats who in client processing redefine their jobs by taking into
account the informal but likely consequences of their actions. Judges and
prosecutors, for example, often make charging and sentencing decisions
based on their expectations of the consequences of subjecting defendants to
the results of sentencing, although formally they are not supposed to con-
sider the quality of correctional institutions in their deliberations. These ten-
dencies have earned some judicial personnel considerable criticism for the
resulting leniency of their approaches. 19 Similar reconceptualizations of the
job were evident among the public housing personnel who, contrary to of-
ficial agency policy, took into account the consequences of placing some
bvored applicants in undesirable housing projects, as previously discussed.
Is escape from bureaucracy desirable? Does it represent a tendency to-
ward responsiveness whose absence is too often deplored? Certainly to the
beneficiaries of these orientations it represents responsiveness. However,
the dilemmas of street-level bureaucracy remain unresolved. Workers who
undermine intake practices by favoring some clients deny minimal services
to those who fail to get entered on the agency rolls. Public housing appli-
 The Client-Processing Mentality
cants who do not receive treatment are disadvantaged because fewer places
are available in the better projects. Judges and prosecutors who develop
private conceptions of proper considerations in charging and sentencing con-
tribute to defendants' welfare as best they can, but they also skew the popu-
lation of the correctional institutions in ways responsive to their private con-
ceptions of appropriate sentencing. And to judge by the proliferation of
mandatory sentencing legislation, they force the development of inflexible
policy to restore the formal order. However one might sympathize with
court personnel who take discretionary actions in clients' interests, one can-
not conclude that they substantially resolve the dilemmas of confining the
scope of discretion and negating the consequences of rule-bound bureau-
cracy.

Modifications of Conceptions of Clients

Street-level bureaucrats are expected to treat all people in common circum-

stances alike. Paradoxically, many factors operate to make favoritism and un-
equal treatment characteristic of modem bureaucracies. These factors in-
clude the inherent subjectivity of required judgments, the difficulty of
assessing street-level bureaucrats' work, the inadequacy offeedback as an in-
fluence on behavior, and ideological considerations that justify client dif-
ferentiation. These concerns have been treated in the previous three
chapters focusing on patterns of practice developed to make jobs easier to
manage.
However, a discussion of the importance of practices resulting in client
differentiation would be incomplete without reference to the psychological
importance of client differentiations as a coping strategy. Client differentia-
tion is a significant aspect of street-level bureaucrats' rationalization of the
contradictions in their work. It is not simply that street-level workers prefer
some clients over others. These preferences also make it possible to perform
flexibly and responsively with a limited segment of the clientele. Thus work-
ers do for some what they are unable to do for all. The street-level bureau-
crat salvages for a portion of the clientele a conception of his or her perfor-
mance relatively consistent with ideal conceptions of the job. Thus as the
work is experienced there is no dissonance between the job as it should be
done and the job as it is done for a portion of the clientele. The worker knows
in a private sense that he or she is capable of doing the job well and can bet-
PATTERNS OF PRACTICE

ter defend against the assaults to the ego which the structure of street-level
work normally delivers. The teacher's pet is not only an obedient child but
also one who confirms to the teacher the teacher's own capability.
There is another important reason to consider street-level bureaucrats'
conceptual modifications of the clientele. Just as differentiation of clients
supports rationing and other practices of organizing work, it also supports
private modifications of conceptions of work. Conceptions of the job imply
conceptions of the clientele. One cannot practice without an implicit model
of the people on whom one is practicing. An open classroom demands a con-
ception of children as requiring relatively greater freedom and flexibility
than are available in a traditional classroom. A psychiatrically oriented drug
center is founded on a different model of human motivation than a center
organized around peer interaction and self-help.
Street-level bureaucrats who are unable to provide all clients with their
best efforts develop conceptual mechanisms to divide up the client popula-
tion and rationalize the division. The differentiation of clients discussed in
previous chapters thus not only provides a rationale for allocating scarce
resources, but it also serves to help street-level bureaucrats justify their jobs
to themselves. The frequency with which street-level bureaucrats are ob-
served to divide up the client world conceptually suggests the importance of
this dimension of work in sustaining street-level practice.
The psychological importance of private reconceptions of the clientele can
be traced in the primary divisions of the client world. For example, unsanc-
tioned distinctions between worthy and unworthy clients narrow the range
of clients for whom street-level bureaucrats must provide their best efforts.
Street-level bureaucrats often respond more favorably to clients who are
helpful or cooperative in their own treatment, or who appear to be particu-
larly responsive to help. Orienting services toward cooperative clients, or
clients who respond to treatment, allows street-level bureaucrats to believe
that they are optimizing their use of resources. At the same time these per-
ceptions help condone service denials (or even routine treatment) by per-
mitting the private judgment that some clients absorb more than their fair
share of resources.
Perhaps the most familiar syndrome of private reconceptions of clients
concerns locating responsibility for client difficulties. Assumptions about
who or what is responsible for clients' situations are significant conceptual in-
struments by which street-level bureaucrats distance themselves from
clients. For example, the tendency of helping professionals to blame the vic-
tim, attributing the cause of clients' situations to the individuals themselves
 The Client-Processing Mentality
without considering the role of social and environmental contexts, locates re-
sponsibility in a place that absolves the helper from blame. 20
There are many examples of blaming the victim. Chronically unemployed
men are described as shiftless and unwilling to work when their situations
might be attributed to the structure of employment and previous job avail-
ability. Students' learning difficulties are explained by focusing on their lack of
motivation rather than on the skills of the teachers and the atmosphere of the
school. Blaming clients for failing to keep appointments protects street-level
bureaucrats from the possibility that prior interviews have discouraged or
alienated them. Instances of teachers beating children who clearly display
signs of mental disturbance provide particularly brutal illustrations of the ap-
parent need of at least some street-level bureaucrats to attribute self-direc-
tion to noncompliant clients. 21 If the client is to blame, street-level bureau-
crats are shielded from having to confront their own failures or the failures of
the agencies for which they work.
An opposite but functionally equivalent mode of perceiving clients also
serves to absolve street-level bureaucrats from responsibility for service fail-
ures. This is the tendency to take an entirely environmental point of view
and perceive clients exclusively as the products of inadequate background
conditioning. Thus if children are perceived as primitive, racially inferior, or
culturally deprived, teachers can hardly fault themselves if their charges fail
to progress. 22 Similarly, job training counselors who explain failures by
clients' low motivation stemming from the discouragement experienced by
ghetto youth can avoid dealing with their own failures to make the program
meaningful.
Undeniably, there are cultural and social factors that affect client perfor-
mance, just as there is a sense in which people are responsible for their ac-
tions. However, it is important to note that these explanations function as
cognitive shields, reducing what responsibility and accountability may exist
in the role expectations of street-level bureaucrats. Moreover, because these
explanations of responsibility are illegitimate in terms of formal agency pol-
icy, they remain beneath the surface, unstated. Thus when they implicitly
form the basis for decisions about clients they contribute to misunder-
standing and to the resulting hostility of clients toward the agencies acting
upon them.
Given the imbalance in power between clients and their agencies, not all
clients will respond with hostility to decisions based on these implicit as-
sumptions. Perhaps more commonly, clients accept the implicit assumptions
of responsibility; then these conceptual structures contribute to client com-

153
PATTERNS OF PRACTICE

pliance with agency policy. Clients may accept responsibility for their cir-
cumstances without reference to the environmental conditions that they ex-
perience. Or they may regard their situation as hopeless because their
environment is so antagonistic to improvement. Each attitudinal set works
against personal movement and growth. 2 3
This is not to say that one can easily strike out for one explanation of re-
sponsibility over another. Structural explanations of clients' circumstances
are important in order to direct attention to changing the political, eco-
nomic, and social structures that circumscribe and dictate the possibilities of
action. For if environmental factors do not prescribe life changes they cer-
tainly structure the range of opportunities.
Similarly, in important respects clients to some degree must be responsi-
ble for themselves. Without this assumption there can be no client growth
within the current structure of arrangements and no client contribution to
changing those arrangements, individually or collectively. Erving Coffman's
insight into the relationship of client responsibility to absolving explanations,
developed in his study of prisons, mental hospitals, and other "total" institu-
tions, has generally wider applicability.
Although there is a psychiatric \iew of mental disorder and an environmental Yiew of
crime and counterreYolutionary actiYity, both tieeing the offender from moral re-
sponsibility for his oflense, total institutions can little afiord this particular kind of de-
terminism. Inmates must be caused to self-direct themselves in a manageable way,
and, for this to be promoted, both desired and undesired conduct must be defined as
springing fiom the personal will and character of the indi,idual inmate himself: and
defined as something he himself can do something about. 24

These views of social responsibility do not originate with street-level bu-

reaucrats, of course. But they are adopted and rigidly held by workers faced
with the contradiction that they ought to be able to make a difference in
clients' lives, but commonly cannot. These views explain failure away, and
permit workers to develop more comfortable relations with the contra-
dictions in their work.
Not all street-level bureaucrats develop these attitudinal patterns. Con-
spicuously, some public services develop different patterns of attribution of
client responsibilities from others, and variations can also be found within in-
dividual public services. 25 The task of those interested in promoting the
quality of street-level bureaucracy is to help sustain the ambiguity in allocat-
ing responsibility. It is undoubtedly an important measure of street-level bu-
reaucratic services that some workers find a way to keep in balance their
views of client responsibility and environmental causality and their own po-
tential for intervention.

154
 The Client-Processing Mentality
Street-level bureaucrats hold private views that affect the distribution and
quality of services, and they hold these views intensely. Their biases, when
they exist, are difficult to interrupt. Why should this be so when street-level
bureaucrats, more than most people, have regular opportunities to discon-
firm stereotypes?
A partial possible explanation has already been suggested. First, segmen-
tation of the client population complements work practices that are them-
selves compromises, and it also complements the resulting reconceptions of
work objectives. In other words, patterns of practice, conceptions of the job,
and conceptions of the clients must fit together if street-level bureaucrats are
to resolve work contradictions successfully. Private conceptions of the clien-
tele will be developed in proportion to the need to come to a private resolu-
tion of the contradictions in the work.
Second, conceptual modifications of the clientele tend to accept and build
upon general social attitudes, and thus are reinforced in everyday life. Fa-
voring clients who are underdogs or discriminating against clients consid-
ered socially unworthy may partly be explained by the sympathies and an-
tipathies of the general society. Sociologist Howard Becker reports that
children may be morally unacceptable to teachers in terms of values cen-
tered around health and cleanliness, sex and aggression, ambition and work,
and age-group relations. These considerations are particularly likely to be
salient when class discrepancies between teachers and pupils are signifi-
cant. 26 These responses to childrens' characteristics are not likely to be
unique to teachers. But when teachers do respond to children in these
terms, their responses have implications for public policy.
Other conceptions of clients appear to enhance feelings about job ac-
complishments even when they seem to run counter to prevailing social
norms. Consider the case of social service workers who would rather be as-
signed to child abuse than to child neglect cases. Although child abuse is a
particularly unattractive crime the anomaly appears to be explained by the
greater likelihood that child abuse cases will respond to intervention, while
the typically passive child neglector is less likely to respond to social work-
ers' assistance. 27 It would appear that clientele segmentation is usually con-
sistent with prevailing social norms, but it is not wholly explained by them.
Third, various aspects of the ways in which street-level bureaucrats re-
ceive information about their work contribute to conceptual modifications of
the clientele. Illustrative validation, self-fulfilling prophecies, rational-
izations that excuse failure, and selective retention of information tend to
confirm rather than disconfirm workers' attitudes about clients.
Finally, street-level bureaucrats work in a milieu in which their co-

155
PATTERNS OF PRACTICE

workers have similar needs to segment the client population. Thus attitudes
prejudicial or beneficial to certain clients are likely to reverberate among,
rather than be contradicted by, other workers.
Street-level bureaucrats have a need to modify their conceptions of clients
quite apart from but usually consistent with the prejudices of the general so-
ciety. And they work in a structure that tends to confirm the validity of their
biases. The general argument of this section, based on observations that
street-level bureaucrats consistently introduce unsanctioned biases into
client processing, suggests that it would be difficult to eliminate client dif-
ferentiation without changing the structure of work for which these biases
are functional.
This is not to say that any particular bias is necessary to cope with the
work. No doubt classes of clients may be treated in markedly different ways
if administrators pay enough attention to specific behavior of workers. But
without changes in the work structure one ought to expect that biases will
soon develop in other areas, or that the old biases will soon emerge in new
forms in the absence of considerable vigilance.
 PART IV

THE FUTURE OF
STREET- LEVEL
BUREAUCRACY
 CHAPTER 11

The Assault on Human Services:

Bureaucratic Control)
Accountability)
and the Fiscal Crisis

This chapter examines the current application of administrative measures to

secure accountability among street-level bureaucrats. I argue that bureau-
cratic accountability is virtually impossible to achieve among lower-level
workers who exercise high degrees of discretion, at least where qualitative
aspects of the work are involved. Nonetheless, public managers are pres-
sured to secure or improve workers' accountability through manipulation of
incentives and other aspects of job structure immediately available to them.
When considered along with other objectives public managers seek, the
results may not simply be ineffective but may also lead to an erosion of ser-
vice quality.
Current perceptions of fiscal crisis heighten concerns for bureaucratic ac-
countability. State and local politicians seeking to cut or constrain budgets
must look to street-level bureaucracies if they want to reduce public
payrolls. As schools, welfare offices, and police departments confront de-
mands for personnel cuts, the issue of accountability arises with urgency. If
public workers cannot demonstrate accountability, all the more reason to
slash their numbers. And if their numbers are reduced, all the more reason

159
THE FUTURE

that politicians seek ways to hold the remainder accountable so as to assure

clients and the public that essential government tasks will continue to be
performed responsibly.
Accountability is the link between bureaucracy and democracy. Modern
democracy depends on the accountability of bureaucracies to carry out de-
clared policy and otherwise administer the ongoing structures of govern-
mentally determined opportunity and regulation. If this is the case, account-
ability must mean more than simply having to answer to a superior, or
expecting to be called to account for one's actions. This is because these defi-
nitions leave open precisely what is at issue: whether there is any reliable
relationship between what superiors seek and what subordinates do.
Taking a different tack, I propose that people should be considered ac-
countable when there is a high probability that they will be responsive to le-
gitimate authority or influence. This definition of accountability directs at-
tention to two important aspects of the concept. First, accountability is a
relationship between people or groups. One is always accountable to some-
one, accountability is not abstract. Although the term is sometimes used
loosely confusion results unless we specifY both parties in the accountability
relationship.l
Second, accountability refers to patterns of behavior. Only if a pattern of
behavior exists can predictability, and therefore accountability, exist. In
practical terms this means that efforts to change or improve accountability
cannot succeed unless patterns of behavior change or improve. For example,
medical and police review boards do not increase accountability unless gen-
eral relationships with citizens or superiors change. This is no more than say-
ing that laws are only effective if they not only punish transgressions but also
deter illegal behavior.
From this perspective attempts to increase accountability through ad-
ministrative controls may be seen as efforts to increase the congruence be-
tween worker behavior and the policies of agency executives through the use
of sanctions and incentives available to the organization. Manipulation of ad-
ministrative controls is not the only way to secure accountability. Recent ef-
forts and speculation have also focused on improving accountability to con-
sumers by recreating the conditions of a market (for example, voucher
proposals), to the public by changing the structure of government programs
(for example, school decentralization), and to the law by seeking judicial
relief. Some emphasis has also been placed recently on improving account-
ability to professional norms by enhancing employees' status and training.
Yet of all these efforts to obtain bureaucratic accountability represent the
range of actions available in the short run to people who manage public agen-

160
 The Assault on Human Services
cies. Whatever theoretical advantages exist in other approaches, efforts to
obtain bureaucratic accountability will have the greatest immediate impact
on workers and clients alike.
To utilize organizational incentives and sanctions, at least the following
conditions must prevail. These conditions are the prerequisites of a bureau-
cratic accountability policy.
1. Agencies must know what they want workers to do. Where the objectives are
multiple and conflicting, agencies must be able to rank their preferences.
z. Agencies must know how to measure workers performance.
3 Agencies must be able to compare workers to one another to establish a stan-
dard for judgment.
4 Agencies must have incentives and sanctions capable of disciplining workers.
They must be able to prevail over other incentives and sanctions that may
operate.

The preconditions of an accountability policy may exist in many bureau-

cratic contexts, but they do not apply where street-level bureaucrats are con-
cerned. Efforts to improve bureaucratic accountability policies in these con-
texts may undermine rather than enhance service quality and may
systematically decrease service quality when certain conditions of public bu-
reaucracy prevail.
The essence of street-level bureaucracies is that they require people to
make decisions about other people. Street-level bureaucrats have discretion
because the nature of service provision calls for human judgment that cannot
be programmed and for which machines cannot substitute. Street-level bu-
reaucrats have responsibility for making unique and fully appropriate re-
sponses to individual clients and their situations. It is the nature of what we
call human services that the unique aspects of people and their situations
will be apprehended by public service workers and translated into courses of
action responsive to each case within (more or less broad) limits imposed by
their agencies. They will not, in fact, dispose of every case in unique fashion.
The limitations on possible responses are often circumscribed, for example,
by the prevailing statutory provisions of the law or the categories of services
to which recipients can be assigned. However, street-level bureaucrats still
have the responsibility at least to be open to the possibility that each client
presents special circumstances and opportunities that may require fresh
thinking and flexible action.
If this is the case, street-level bureaucrats must irreducibly be accountable
to the client and to an appropriate response to the client's situation and cir-
cumstances. These considerations cannot sensibly be translated into authori-
tative agency guidelines, although it is on behalf of their agencies that street-
THE FUTURE

level bureaucrats are accountable to clients. It is a contradiction in terms to

say that the worker should be accountable to respond to each client in the
unique fashion appropriate to the presenting case. For no accountability can
exist if the agency does not know what response it prefers, and it cannot as-
sert a preferred response if each worker should be open to the possibility
that unique and fresh responses are appropriate. It is more useful to suggest
that street-level bureaucrats are ordinarily expected to be accountable to two
sources of influence-agency preferences and clients' claims. 2
There are other sources for the assertion that street-level bureaucrats are
ordinarily expected to be accountable to clients in possible conflict with the
agencies for which they work. The most important of these is that most
street-level bureaucrats are professionals or they work in occupations aspir-
ing to professional status. In either case a fundamental expectation attached
to the job is that client needs are primary and that the extension of public
trust depends upon accountability to people as individuals when they are en-
countered in the course of work. Social workers, teachers, and, of course,
doctors and lawyers, are expected to respond to the individual and the
presenting situation, however strongly their work situations mitigate against
flexible responses.
This is a great strength and also a great weakness of the public services. It
provides a measure of responsiveness to clients when the organization of bu-
reaucratic service tends toward neglect or rigidity. But by virtue of provid-
ing another focus of accountability it also means that street-level bureaucrats
are less controllable.

Holding Workers to Agency Objectives

Despite the dual focus of accountability inherent in the street-level bureau-

crats' roles, public managers are drawn to making street-level bureaucrats
more accountable by reducing their discretion and constraining their alter-
natives. They write manuals to cover contingencies. They audit the perfor-
mance of workers to provide retrospective sanctions in anticipation of which
it is hoped future behavior will be modified. They insist workers specify ob-
jectives in the hopes that accountability can be more effectively monitored.
These management tools at times may be effective in controlling workers.
Manuals specifying proper procedures may help standardize responses and
provide instruction ..Performance audits may create greater awareness that
 The Assault on Human Services
management is observing performance, and may thus lead workers to take
greater care. Specifying objectives is always likely to be instructive and it
directs workers' attention to the relationship between the available re-
sources and the goals they are trying to achieve. 3
However, street-level bureaucrats may subvert efforts to control them
more effectively in the name of accountability. In these and other examples
of attempts to increase control it is relatively easy for workers to tailor their
behavior to avoid accountability. For one thing, they are likely to be the
source of information management receives concerning their performance.
They are fully able to provide information about the presenting situation that
makes the action taken appear to be responsive to the original problem when
it may not have been. This is less blatant falsification than it is auspicious
shading of the truth and sincere rationalization.
It is extremely difficult for management to contradict workers' reports for
several reasons. A critical piece of information is the state of mind of the
worker and his or her analysis of the presenting situation. Since street-level
decisions are made in private it is extremely difficult to second-guess work-
ers, since the second-guessers are not at hand to evaluate the intangible fac-
tors that may have contributed to the original judgment. For this reason, the
records kept by street-level bureaucrats are almost never complete or ade-
quate to the task of post hoc auditing, and when records are kept, they are
written sketchily and defensively to guard against later adverse scrutiny. 4
Record keeping can help insure that procedures are followed (since fal-
sification is normally not the issue). Health practitioners can be made to run
certain tests, social workers to ask certain questions, police officers to follow
certain procedures. But the records cannot force accountability on the ap-
propriateness of the actions to the presenting situation.
Another major difficulty with obtaining accountability through manage-
ment control arises because of the dependence of street-level bureaucracies
on their workers. Since the services delivered by schools, police depart-
ments, or legal services offices fundamentally consist of the actions of
teachers, police officers, and lawyers, these agencies are constrained from
controlling workers too much, particularly in challenging their performance,
for fear of generating opposition to management policies and diminishing ac-
countability even further. The weakness of management incentives to sanc-
tion negative performance contributes to a climate in which vigorous chal-
lenges to street-level bureaucrats' autonomy in decision making is presumed
to have possible negative net consequences for service delivery, by destroy-
ing morale and inhibiting worker initiative. 5
Are there negative aspects to management control efforts, or are these ef-
THE FUTURE

forts simply generally ineffective? There are several respects in which con-
trol practices can actively subvert service quality.
First, specification of methods of client treatment under the guise of ob-
taining accountability may actually result in reductions in client services.
There is often a thin line between inducing workers to better conform to
agency policies and inducing workers to be open to fewer options and oppor-
tunities for clients. For example, during the Nixon Administration the De-
partment of Health, Education, and Welfare attempted to increase welfare
employees' accountability by auditing their error rate in accepting clients for
welfare. This policy reduced services by providing incentives for welfare
workers only to reduce errors that favored clients. Federal guidelines did
not then call for reducing error rates for the potential welfare population as a
whole. If it had done so the applications of all welfare applicants would have
been audited, both those accepted and those rejected. Scrutiny of welfare
workers' decisions strictly in terms of whether or not they were too lenient
amounts to narrowing the role of welfare workers, reducing their account-
ability to clients and to professional standards of conduct.
Second, supervision of subordinates with broad discretion and responsi-
bilities requires assertions of priorities in attempting to increase account-
ability. Police departments may scrutinize traffic tickets, vice arrests, or in-
terracial encounters between police and citizens. But they cannot
meaningfully hold officers accountable for everything all the time. If every-
thing is scrutinized, nothing is scrutinized. Thus efforts to control street-
level bureaucrats not only affect those areas that are management targets,
but they also affect those areas that are not the focus of management efforts,
since by implication those efforts will not come up for surveillance. Efforts to
increase accountability in some areas may come to be regarded as the only
areas in which accountability will be sought and behavior scrutinized.
Third, many management control efforts provide a veneer of account-
ability without in fact constraining behavior very much. Management con-
trol systems have symbolic value, providing concerned publics with reassur-
ances that employees are accountable even when they are not. Introduction
of management systems at least temporarily permits agencies to deflect criti-
cisms as citizens find it very difficult to challenge the emperor who officials
say is fully clothed, appearances and personal experiences to the contrary
notwithstanding. 6

GOAL CLARIFICATION
One of the conspicuous features of many public services is the ambiguity
and multiplicity of objectives. How can accountability be achieved, ask the
 The Assault on Human Services
critics, if public officials are unclear about their objectives? The desirability
of clarifying (and then putting into operation) agency objectives to increase
accountability stems from the force of this observation and recognition that
a bureaucratic accountability policy requires specification of objectives (as
suggested above).
Surely it is desirable to clarify objectives if they are needlessly and irrele-
vantly fuzzy or cgntradictory. Surely it is easier to run an effective agency if
you know what you are supposed to be doing. However, while agency goals
may be unclear or contradictory for reasons of neglect and historical inertia,
they may also be unclear or contradictory because they reflect the contra-
dictory impulses of the society the agency serves. Schools attempt to in-
struct, but they also inculcate attitudes toward social behavior and c'itizen.-
ship. They do this not because educators are fuzzy but because both these
objectives are favored by parents (and because there is no convincing case
that they are mutually incompatible). Criminal justice institutions are
oriented toward punishment and rehabilitation not because judges and cor-
rections officials are simpleminded but because the society has impulses
toward reforming as well as deterring criminals.
The public service areas of education, corrections, and welfare in recent
years have all been subject to efforts to increase accountability through goal
clarification. Educators have sought to concentrate on reading to the exclu-
sion of other educational objectives; corrections analysts have sought to build
up the role of punishment and make it more certain at the expense of empha-
sis on rehabilitation; welfare reformers have successfully separated decisions
on income support from social service provision. The dilemma for account-
ability is to know when goal clarification is desirable because continued am-
bivalence and contradiction are unproductive, and when it will result in a
reduction in the scope and mission of public services. The problem of goal
ambiguity has contributed to the discrediting of institutions providing ser-
vices in social work, corrections, and mental health, and to the dismantling
of many programs to provide assistance in these areas. But it requires the
most serious inquiry to determine the long-term implications of requiring
the former and potential clients of these institutions to have recourse exclu-
sively to noninstitutional community and personal resources.

PERFORMANCE MEASURES
The development of performance measures is critical to a bureaucratic ac-
countability policy. Administrators make great efforts to develop perfor-
mance measures in order to control employees' behavior.
THE FUTURE

There is no question that public services can be enhanced through devel-

opment of valid performance measures. In such cases public service workers
can be held accountable for producing results in the same way that machine
operators can be charged with producing a certain volume of output in a
given period of time. However, public service workers, like machine opera-
tors, must also be assessed for quality control, since producing a volume of
items is meaningless without consideration of the standard maintained in
production. Here, paradoxically the search for performance measures can in-
terfere with the quality of public service.
In theory quantitative measures of performance should be fairly easy to
obtain and consent on their validity reasonably uncontroversial. This is not
always the case in street-level bureaucracies, however, for several reasons.
First, street-level bureaucrats will concentrate on the activities measured.
If police officers are assessed on traffic ticketing or vice arrests, activity in
these areas will increase. This is entirely predictable when we recognize that
police have control over their search activities and can choose to concentrate
on one dimension of their job or another. By virtue of simply putting atten-
tion on some tasks over others street-level bureaucrats can improve their
performance on most quantitative measures managers introduce. If welfare
workers are assessed on their error rate, the error rate will go down because
workers pay more attention to it. If teachers are assessed or even remotely
evaluated on the proportion of their charges who pass year-end examina-
tions, more will pass as teachers "teach the test." This is neither surprising
nor in itself deplorable, but simply highly probable. Whenever management
undertakes to concentrate on measuring a dimension of performance work-
ers correctly accept this as a signal of management priority. A problem is
created, however, when the measure induces workers to reduce attention to
other aspects of their jobs and when there is no control on the quality of
work produced. 7
Relatedly, street-level bureaucrats will make choices and exercise discre-
tion by directing their activities in ways that will improve their perfor-
mance scores. This phenomenon did not begin and end with Peter Blau's
classic-report of the employment counselors who, when assessed in terms of
successful placement ratios rather than the case load they carried, made
greater efforts for easy-to-place clients at the expense of more difficult
cases. 8 The phenomenon of"creaming" in recruiting for social programs has
similar dynamics. Workers select for their programs clients likely to do well
in them, in order to improve the appearance of success. As James D.
Thompson says, "[W]here work loads exceed capacity and the individual has
166
 The Assault on Human Services
options, he is tempted to select tasks which promise to enhance his scores on
assessment criteria." 9 This generalization obtains for individuals and also for
the work units of which they are a part.
Fraud and deception can also intrude into performance measurement.
The Washington, D.C. police took pride in their record ofreducing serious
crimes until a study revealed that police officers were reporting that most
burglaries involved items valued at less than fifty dollars. Significantly, the
definition of a serious (that is, felonious) burglary is defined in part as involv-
ing the theft of over fifty dollars in value. 10 The incentives to under-report
the value of items in burglaries are the same as those that induced New York
City sanitation men to water their garbage so their trucks would weigh the
expected amount when they appeared at the landfill site, even though the
drivers had not completed their runs.
It may be claimed that these problems-inducing behavior to conform to
the measure, neglecting other responsibilities, and inauthentically perform-
ing according to the measured standards-are simply difficulties that skilled
management experts can overcome. In particular, management often seeks
measures of resource deployment, depending on the inference that the pro-
vision of resources is a surrogate for (and, to be sure, a prerequisite to) ser-
vice delivery. This inference is acceptable when the qualitative issue is
resource deployment, as in the case of police dispatch, ambulance response
time, and neighborhood shift allocations in sanitation services. 11
The difficulty arises in the inference that resource deployment of a partic-
ular sort bears a relationship to the quality of service delivery. For example,
case-load activity might be used as a quantitative measure of performance,
since it indicates formal relationships between street-level bureaucrats and
clients. Class size indicates associations between teachers and children.
Court dispositions indicate relationships between defendants and judicial
personnel. But in all these instances there may be inverse relationships be-
tween the quality of street-level bureaucrats' involvement and the number
of clients they process. If simply having people processed or having them at-
tached to public service workers were the issue, these measures would bear
a meaningful relationship to desired service. But our expectations of these
public services are different. It is not sufficient that people are assigned a
social worker, sit in a classroom, or have their cases heard. We also expect
that they will be processed with a degree of care, with attention to their cir-
cumstances and potential. Thus there may be no relationship, or an inverse
relationship, between quantitative indicators of service and service quality.
The more discretion is part of the bureaucratic role, the less one can infer
 101:. .t'UlUfll:.

that quantitative indicators bear relationship to service quality. Even in such

an apparently straightforward measure as the number of arrests made by
police officers or the number of people treated in emergency rooms, we have
no idea whether the arrests were made with care or that treatment met ap-
propriate standards. Sophisticated _!Ilanagement specialists acknowledge the
problems of inferring quality from quantitative indicators. 12 But this does
not prevent them from using quantitative measures as surrogates for service
quality or ignoring the problems of inference in their utilization. 13
Of course the reason quantitative measures are used so dften is that actual
performance is virtually impossible to measure. It is perhaps useful to put
this quite bluntly. We cannot measure the quality of street-level bureau-
crats' performances, particularly in terms of the most important aspects of
their jobs. Aspects of performance can be measured and assessed, and many
surrogates for performance measures can be developed with important im-
plications as management tools. But the most important dimensions of ser-
vice performance defy calibration.
Measures of performance quality are elusive for reasons analogous to the
difficulty in circumscribing street-level bureaucrats' discretion. If clients or
presenting cases should be treated as if they might present unique situa-
tions, then it is impossible to reduce responses to sets of appropriate and
previously indicated reactions. To put it another way, the more street-level
bureaucrats are supposed to act with discretion, and the broader the areas of
discretionary treatment, the more difficult it is to develop performance mea-
sures. If we are not agreed as to what comprises good teaching, how can we
measure it? If we are not willing to deprive police officers of discretion
because on the street they need to be able to make judgments based on an
appraisal of the total situation, how can we propose measures of quality ar-
rests and interventions with citizens? If every client should be treated as if
he or she may require responses tailored to the individual, how can we spec-
ify the requirements of a good interview?
It may be argued that we may still assess service quality by developing
outcome indicators. But here similar questions arise. First, service quality
measures are meaningless without adequate controls to assess levels of dif-
ficulty. The same outcome may have required radically different service
because of the difficulties presented. For example, the same student
achievement levels might represent excellent work on the part of a teacher
of students with learning difficulties and poor work on the part of a teacher
with bright and motivated students.
Without controls there can be no comparability of units of analysis, unless
the often unwarranted assumption is made that levels of difficulty are equal.

168
 The Assault on Human Services
Thus teachers resist being measured by the progress of their pupils unless
adequate provision is made to control for their students' previous levels of
achievement (and, more important) for their students' capacity to learn.
Thus police officers would object to utilizing arrests per capita per available
officer as measure of performance unless controls were introduced for the
propensity of criminal behavior in the district. Comparing districts by out-
come tends to be useless because of the inadequacy of such controls.
Some advocate that measures such as these be deployed in order to dis-
cover deviations from normal practice, so that workers who deviate from the
norm can be brought into line. These proposals have some merit, but here
the problem is that unless one is confident that the best workers or dis-
tricts are doing a good job, such comparisons may simply institutionalize
mediocrity.
Recent attention has also focused on utilizing surveys of client satisfaction
to obtain information on workers' performance. Resistance to such surveys
arises out of professional skepticism that performance that pleases clients
may not be related to high-quality practice. The popular teacher may not be
the most effective instructor; a popular judge may not be the fairest judge.
Still, in a comprehensive assessment program, client satisfaction surveys
have their place if client satisfaction is indeed deemed desirable (as it usually
is).
Street-level bureaucrats' interactions with clients tend to take place in
private or beyond the scrutiny of supervisors. Interviews are held in private
offices and under norms of confidentiality. Teaching is done in classrooms
that principals and supervisors do not normally enter; if they do, they provide
notice so that the teaching, like a performance, may be changed by the pres-
ence of an audience. Police officers, although they do take action in public,
normally do so without observation by other officers or supervisors. The ex-
ception is the officer's partner, who is compelled by police norms to shield
his partner from criticism. Of the street-level bureaucrats we have studied
only judges tend to make their important decisions in public.
This fact provides a barrier to an important potential source of perfor-
mance measurement. It might be possible for street-level bureaucrats to
scrutinize each others' work and provide assessments of quality. But given
the structure of these agencies such scrutiny would be highly obtrusive in
relations between workers and clients and very costly if engaged in on a
widespread basis. Thus public service agencies rarely engage in direct obser-
vation of their line workers, but depend upon the written record supplied by
their workers (the reliability of which was discussed earlier).

169
 Accountability and Productivity

Thus far I have focused discussion on some of the major difficulties in devel-
oping an administrative accountability policy. But are there any negative ef-
fects of such policies? For example, what is the harm of attempting to de-
velop performance measures? It may be difficult to measure performance,
but perhaps we are simply at the beginning of the development of a manage-
ment tool. Perhaps the current measures of performance are not entirely ad-
equate, but they may have their uses, and they may be increasingly refined.
This is the rhetoric of those who are committed to achieving bureaucratic
accountability and recognize the inadequacy of current measures, but who
apparently have faith that their approach is ultimately oorrect. 14 This line of
discussion would have us believe that there are some benefits to current ef-
forts to develop accountability through improved performance measure-
ment, that these benefits are likely to increase, and that there are no signifi-
cant costs. Surely management benefits from operationalizing and
attempting to develop measures of worker performance. Even if the pre-
ferred behavior cannot be adequately measured, performance measurement
and monitoring can signal workers powerfully concerning which aspects of
performance are most salient.
However, in the current period bureaucratic accountability policies also
have negative consequences because of the competing demands on, and of,
administrators. Currently, public agencies are under enormous pressures to
minimize costs and increase productivity. They are under pressure to reduce
government expenditures or keep them from rising. They are under pres-
sure to increase productivity in order to maintain services, or claim that they
are maintaining them, in the face of financial stringency. And they are under
pressure to increase productivity in order to justify employee pay increases,
which they are under particular pressure to grant because of the impact of
inflation on wages. (The only way to stabilize government budgets when ser-
vices cannot be reduced beyond a certain level and costs are rising is to
increase the productivity of the present work force. Organized workers
argue 'that they have no incentive to increase productivity unless they can
share in the gains made because they work harder and cooperate with the
reorganization of work often entailed by productivity reforms.) 15
Productivity in the public sector summarizes the relationship between the
utilization of resources and the resulting public services product. Productiv-
ity may improve in three ways: when costs remain the same while public ser-
vices increase; when costs decline while services remain the same; or when
 The Assault on Human Services
costs increase but services increase still more. Schematically, there are two
dimensions to public services implied here-<>ne qualitative, the other
quantitative. If the quantity increases or remains the same, but services have
declined qualitatively, productivity increases have not necessarily taken
place. 16 If more garbage is picked up on the streets without increasing the
crews, but half of it is strewn back on the streets, productivity has not
increased. The debasement of service is what infuriated New Yorkers when
transit workers were given an increase in pay based upon alleged gains in
productivity. It appeared, however, that the Transit Authority had been able
to provide services with fewer personnel only by increasing the time be-
tween trains and reducing the number of cars in operation. 17 In this view the
transit workers had been falsely credited with improved productivity.
These are the essential elements of productivity. In practice, however,
debasement of services is rarely taken into account, although the problem is
given lip service by productivity theorists. This may be true for several
reasons. First, if the quality of service is difficult to measure, so is reduction
in service quality difficult to measure. Second, there are many ways to save
money by eroding the quality of service without appearing to do so. They
include offering services on a group rather than an individual basis, substi-
tuting paraprofessionals, often paid from other sources, for regular staff, and,
conversely, forcing professionals to handle clerical and other routine chores,
reducing the time they have to interact with their clients. 18 Additionally,
street-level bureaucrats can narrow the range of situations in which they will
act. Examples include legal services offices that decide to take only
emergency cases, police departments that decide to neglect selected infrac-
tions, and schools that offer reduced programs. Each of these techniques
permit managers tp give the appearance of maintaining services while reduc-
ing costs.
Third, in the current period pressures experienced by public managers to
reduce the budget and improve productivity are pressing and general, while
the constituency for maintaining service quality is disorganized, weak, or
nonexistent. Only clients experience service quality reduction, and they are
severely constrained in comparing their experiences with others and organ-
izing collectively to oppose service quality debasement. Ironically, the
greatest opponents of service quality debasement are street-level bureau-
crats themselves. For them debasement often means harder work, less job
satisfaction, and greater individual problems with clients. Yet they are cross-
pressured by the interest they have in helping their agencies appear finan-
cially responsible and by their collusion with public officials to share finan-
cially in productivity increases.
THE FUTURE

But there is more to the debasement of public services than pressures and
interests. A large part of the problem stems from the orientation toward
measurement, precision, and scientific management itself. Consider the
formula productivity =service quantity and quality/cost. Two of the deter-
minants of productivity, service quantity and cost, are easy to measure; the
third, service quality, is virtually impossible to measure. Managers under
pressure to improve productivity are likely to try to cut personnel or obtain
more work from existing personnel because these are the terms of the eq ua-
tion for which measures are available and which managers can manipulate.
Thus staffs are reduced to bare bones without reduction in responsibilities.
Thus staffs are asked to do more without increases in personnel. 19

Street-Level Bureaucrats and the Fiscal Crisis

There is always an implicit tension between resource constraints and the in-
exorable demands for public service. However, this tension is rarely mani-
fest politically. The budgets (and employment rolls) of street-level bureau-
cracies rise not only with increases in population to be served, but also with
higher standards of what citizens are entitled to, a decline in the availability of
comparable private services, and the perceived need for more effective and
improved agencies of social control, coercive or manipulative. The impulses
to increase expenditures in these areas rarely have been challenged in terms
of resource constraints. In the period since World War II federal govern-
ment subsidies to state and local public services have postponed or softened
the confrontation between revenues and expenditures in areas such as public
health, education, police, and welfare. However, the current period, charac-
terized as a fiscal crisis among state and local governments, forces recogni-
tion of the relationship between what people get from government and what
jurisdictions are willing to pay.
At best the term "fiscal crisis" is reserved for situations in which financial
agreements and long-standing patterns of practice can no longer be honored,
as when a political jurisdiction cannot meet its payroll or honor its commit-
ments to lenders. But the term is also used much more loosely to mobilize
people to believe that there is something wrong or that there is a problem as-
sociated with current and projected expenditures relative to available reve-
 The Assault on Human Services
nues and other income. If political and economic elites are successful in
promulgating a sense of crisis they are able to make manifest and set the
terms of confrontation between governmental expenditures and income. If
in other times social services, for example, grew in response to perceived
societal needs, in a fiscal crisis the imperatives for service development are
subordinated to the demands of perceived revenue limitations. 20
Like other political confrontations, the management of fiscal crises has re-
distributive consequences. The costs of responding to the needs of expendi-
ture constraints do not fall evenly or randomly on the population as a whole,
but rather affect different segments of the population differentially. The fis-
cal crises of the cities provide a focus and an apparently benign rationale for
attacking and injuring the provision of public services. And they demon-
strate the vulnerability to attack of high levels of public service quality.
The fiscal crisis of the cities affects the quality of service delivery in two
significant ways. First, services are rationed in various ways, maintaining the
appearance of service while reducing and debasing it in practice. This is not
to say that legitimate savings cannot be realized by eliminating "real" waste
and duplication that may exist. 21 However, in city administration these
"real" savings tend to be concentrated in areas in which questions of re-
source deployment are paramount, not in areas in which the provision and
nurturing of interactions with street-level bureaucrats is at issue. It is con-
spicuous, for example, that when administrators take pride in productivity
savings it is in the sanitation department that the greatest successes are often
realized. In police departments dispatch (deployment), not interactions with
citizens, is the area of concentration in productivity campaigns. Most street-
level bureaucracies have to contend with impulses to reduce the amount of
time workers spend with clients, not the reverse.
For public officials the problem of managing the fiscal crisis consists of
reducing expenditures while minimizing the apparent impact of the cuts.
This is why cuts will initially be said to eliminate waste and duplication
whether or not waste exists or duplication takes place. Rationing typically
means increasing the costs to clients of seeking services, while maintaining
the service shell, or reducing services to decrease potential benefits. Both
prospects are likely to lead to lower client demand. Closing neighborhood
branch offices while continuing to offer services from a central (downtown)
location is a typical technique for achieving this. Reducing the number of
telephones or receptionists reduces the number of inquiries. Increasing the
response time in investigating a complaint 1educes the efficacy of complain-
ing and hence future volume. City agencies often experience decreasing

173
THE FUTURE

ability to process citizen needs at the same time that they are experimenting
with techniques to improve their performance. The public message regard-
ing agency responsiveness becomes mixed, to say the least.
When public managers decide to fight demands for service reduction they
say that all waste, duplication, and nonessential services have been cut, and
that any further cuts will be in essential services. Their ability to make this
claim depends on general public perception of the importance of the agency
and public employees' collective capacity to resist. Thus schools can make
the claim more effectively than welfare agencies, police departments more
effectively than sanitation departments. Cuts in service provision obviously
may affect service quality, but is is impossible to determine from public rhet-
oric where the politics of distributing urban resources to public employees
ends and injury to service delivery begins.
A second dimension of response to the fiscal crisis is personnel reductions.
Personnel practices are particularly important because salaries comprise the
bulk of urban budgets; thus savings must be sought in the area of public
employment.
Personnel practices in the fiscal crisis tend to follow a path of severity. Ad-
ministrators first make it more difficult for agencies to replace workers who
leave. Next administrators suspend hiring, then slow wage increases or
freeze wages, then begin to lay off workers. All of these steps have implica-
tions for service provision but the most important point is that each of these
steps reflects increasing penalties to public employees, regardless of the
implications for clients. Priorities are set by imperatives of labor-
management relations, not by the needs of clients for service provision.
Increasing the difficulty of replacing workers and freezing employee rolls
represent managers' efforts to realize savings through attrition. They at-
tempt to reduce personnel rolls by not replacing those people who retire or
otherwise leave their jobs. Since in normal circumstances the rate of exit in
almost any line of work is substantial, a significant reduction in employment
can be realized over several years without firing anyone. Realizing savings
through attrition accomodates public managers' needs to keep peace with
organized public employees, but it creates substantial costs for service provi-
sion above and beyond the obvious reduction in force. Since workers do not
retire or otherwise leave their jobs in response to agency priorities, the in-
cidence of turnover is unevenly distributed in the work force. This means
that important gaps in service provision occur. Workers important to the
operation oi a particular office, or those who possess critical skills, may be
the ones to quit rather than the most marginal employees. If the critical posi-

174
 The Assault on Human Services
tions they vacate are left unfilled, the injury to service provision is obvious.
But even if they are filled by employees who remain they are unlikely to be
filled well.
In the street-level bureaucracies with a high level of job specialization the
vacancies will be filled by employees who lack the required skills or re-
sources. Fifth-grade teachers will be assigned to kindergarten classes, and
physical education instructors will become math teachers when there are ex-
cesses of the former and need for the latter. But even in street-level bureau-
cracies with low levels of specialization it will be difficult to fill the vacancies
adequately. Police departments, for example, assign desk officers to active
patrol in order to maintain the patrol force when additional officers cannot be
hired. But those officers assigned to desk jobs may be more suitable for
sedentary positions than the officers they replace. Moreover, except for re-
tirees the people whp leave public service work in the fiscal crisis will tend to
be the more employable, so that the work force at the same time becomes
mediocre through attrition.
The squeeze of the fiscal crisis is particularly tight because in many ways it
reflects national rather than local conditions. If a city were financially
strapped in an otherwise healthy economy, young public employees would
be able to find work in other locations. However, all governments more or
less experience pressure to hold the line on costs at the same time, since the
fiscal crisis is significantly a function of national economic and political
trends. Thus no longer are there jobs waiting for young teachers or social
workers in the suburbs, or California, or elsewhere, as in the past.
Street-level bureaucrats' performance is not tied directly to wage incen-
tives. Promotions and raises, when they are given out, do not depend so
much on job performance as on personal relations, additional outside train-
ing, work-load handling, and other factors unrelated to client servicing.
Moreover, promotions to positions of greater responsibility are rare, since
the job hierarchies in most street-level bureaucracies schematically resem-
ble relatively flat pyramids, with existing jobs undifferentiated at the bottom
of the scale. When most teachers can look forward only to being teachers,
most police officers only to being police officers, incentives to high-quality
service may flag unless specially encouraged. 22
For these reasons it is probably incorrect to argue that wage freezes di-
rectly affect workers' motivation. Their impact is likely to be somewhat dif-
ferent. First, wage freezes and slowdowns affect the likelihood that workers
will stay in their jobs, and they force people out of jobs without firing them.
Again this affects the distribution of age and skills in street-level bureau-

1 75
THE FUTURE

cracies. Older workers will stay to protect and build pension benefits.
Younger workers will be more likely to leave public employment, somewhat
tempered by the availability of jobs elsewhere.
In periods of high inflation wage freezes effectively reduce workers' real
income as well as their income relative to workers in other sectors. The feel-
ings of deprivation that may result from diminishing workers' income in
these ways are quite different from attitudes toward wages that street-level
bureaucrats may have had in other periods. While wages may not play a
major role in creating incentives to performance, effectively reducing wages
may have a significant impact on street-level bureaucrats' attitudes toward
their jobs.
Street-level bureaucrats may be accountable to managers and clients, as
mentioned above. But they are accountable to clients on behalf of their
agencies and the public purposes they represent. To reduce workers' wages is
to shred these bonds of accountability by bringing to the surface aspects of
the wage relationship that otherwise remain obscured by the claims and
ideology of professional status and attitudes. People who accept the rela-
tively fixed civil-service formulas for wage increases will not easily accept
receiving less than they did receive, particularly as the motivation for street-
level work leads some employees to regard their work as voluntaristic to
some degree, that is, undercompensated to begin with. As wages are effec-
tively reduced street-level bureaucrats may be expected to look more to
their benefits and remuneration and less to the service dimensions of their
job.
Decruitment practices, of which firing workers is the most drastic, have
similar implications. 23 When workers are not replaced their responsibilities
are distributed among those who remain, usually without reducing the re-
sponsibilities they already have. Increasing the responsibilities of other staff
without increased compensation or work resources is the white collar equiv-
alent of the onerous assembly line speedup.
In this connection consider the modest complaint of a New York City
school ,teacher over staff reductions, increased responsibilities, and reduced
resources.

I have never been quite sure what ["increased productivity"] means exactly. How-
ever, if it means what I think it means, they would like to see us work harder than we
ever did before. If this is so, then all the proponents of "increased productivity" will
be delighted to know that we are doing remarkably well in that department.
For example, we have official classes of 45 or more youngsters and ten minutes in
which to take attendance, read circulars, distribute notices, make reports (in dupli-
cate yet), answer questions, etc. In many cases we have classes which have rosters of
 The Assault on Human Services
49 or more children with 30 chairs in the room or typing classes of 47 with 3Z type-
writers. Add to this emergency coverages of classes, cafeteria patrol or other building
assignments, program problems, shortage of supplies and equipment and much
more-all of this with reduced staff-not to mention the mounds of work we take
home with us. The pressures under which we work can never be understood by any-
one who is not involved in day-to-day school activity. Yes, we have indeed increased
productivity, but in so doing we have decreased our effectiveness as human beings to
our students, our families and ourselves. 24

Or consider the impact of the budget freeze on human services in Mas-

sachusetts, when staff attrition resulted in greater physical and chemical re-
straint of mental patients in state institutions. As "hundreds and hundreds of
staff quit their jobs, unable to cope with the conditions and trauma imposed
by a freeze which imperiled their physical safety and adequate care for resi-
dents for whom they bore responsibility," 25 programs to make patients more
self-sufficient were entirely abandoned and the institutions went from care
facilities to "warehouses" for the mentally disabled.
When workers are fired, managers also abridge the implicit contract with
employees by bringing to the surface the reality of job insecurity in what
previously had seemed secure employment. Again, those who remain carry
the work load of those who were fired. While some increased and heroic ef-
forts may be forthcoming it is equally likely that work will be processed in
ways which reduce the amount and quality of time street-level bureaucrats
can spend with clients. 26
Again, this decruitment does not take place according to calculations of
impact on service, but in response to the ethics of seniority. For example,
different public agencies are typically asked to reduce their forces by a cer-
tain fixed percentage. Politicians do not want to choose between services and
so establish a decision rule for work-force reduction. Street-level bureau-
cracies that do choose between units or services undermine the strategic
implications of their choices by providing those with seniority the chance to
"bump" less senior employees from their positions in sectors of the service
that are untouched by the budget cuts.
For some enterprises reductions in personnel rolls can be cleansing but
public agencies cannot take advantage of opportunities to eliminate ineffec-
tive workers. Particularly affected by decruitment are younger workers and
those who have been recently hired, often members of minority groups. It is
difficult to specify the impact of decruitment precisely but we can see that it
falls unevenly on those with more recent training and new ideas and those
with fresh perspectives. It falls less heavily on those with more experience,
an arguable benefit. 27

177
THE FUTURE

These observations concerning reductions in work force should not lead us

to conclude that street-level bureaucracies should be the size they are, or
larger, or that they should never be smaller than they presently are. The fact
is that we know very little about the proper size of street-level service deliv-
ery units. A public school deprived of its specialists might actually serve
students better by throwing regular teachers back on their own resources.
The important point is that when public bureaucracies are growing, jobs
usually are created in response to perceived needs, and workers are hired
with reference to credentials and apparent qualifications. But when they are
forced to shrink, public bureaucracies rarely remove workers in response to
decisions concerning the most effective utilization of reduced available
resources.
Who speaks for service quality in the era of performance measurement,
productivity campaigns, and fiscal crisis? Public managers, with better con-
trol over costs and resource deployment than over the quality of the product,
sacrifice service quality in the name of efficiency and productivity. Street-
level bureaucrats more and more are reduced to production units whose
work is speeded up and whose managers appear content to sacrifice quality
in order to maintain volume. In the process the conditions of work are
eroded and workers are unable to utilize many of the coping mechanisms
and attitudes that helped sustain their jobs under difficult conditions in ear-
lier periods. Thus, the fiscal crisis raises the salience of the wage relationship
and diminishes the salience of service. This is ironic since the wage and
benefit demands of organized public employees have been widely regarded
as one of the primary causes of the fiscal crisis in the first place.
The federal role in this aspect of the fiscal crisis is worth noting. Funds
available through counter-cyclical revenue sharing and the Comprehensive
Employment and Training Act (CETA) underwrote city payrolls for several
years. This helped many cities avoid confronting the enormity of their finan-
cial obligations to street-level bureaucrats. However, current policies based
on the view that the federal role in managing the urban fiscal crisis is less im-
portant than limiting the federal budget threaten all at once to reveal the in-
capacity of cities to retain public employees. Restricting CETA funds to the
chronically unemployed may be sound manpower policy, but it threatens to
throw cities' skilled, middle-class, service-providing agencies into disarray.
This chapter has drawn attention to the contributions of bureaucratic ac-
countability policies to exacerbating problems of the quality of service deliv-
ery. But there is more. Such policies set the stage for future management of
service delivery and future conceptions of the role of human services. If cur-
rent administrative practices erode workers' sense of responsibility for
 The Assault on Human Services
clients, then establishing nonmanipulative, responsive worker/client rela-
tionships will be that much harder in the future. If qualitative aspects of ser-
vice delivery are neglected, cost reductions and volume receive more atten-
tion as workers and managers accommodate their behavior to agency signals
of priorities. This contributes to the self-fulfilling prophecy of the ineffec-
tiveness and ultimate irrelevance of social services, even though the human
needs for nurturing, protection, support, and assistance remain unan-
swered. Thus the tones of the fiscal crisis may linger, even if the budgetary
alarms of the current period are eventually quieted.

1 7Y
 CHAPTER 12

The Broader Context of

Bureaucratic Relations

In considering the potential for change in street-level bureaucracies it would

be a mistake to restrict analysis to the coping dilemmas and adaptations of
service workers, or the patterris of practice that develop among them. The
resolution of contradictory tendencies in street-level bureaucracies cannot
be understood without examining the role of these public agencies in the so-
ciety and the ways in which the society impinges on the character of bureau-
cratic relations. As V. 0. Key Jr. has observed: " ... one of the great func-
tions of the bureaucratic organizations is as a conservator of the values of a
culture. In the purposes, procedures, ceremonies, outlook, and habits of the
bureaucracy are formalized the traditional cultural values. " 1 This observation
actively translates into reciprocity between the larger society and the struc-
ture of bureaucratic institutions. For street-level bureaucracy it means that
these agencies are embedded in a larger system that creates and fortifies
working conditions. In turn, street-level bureaucracies help reproduce pre-
vailing relations between individuals and government organizations. 2
Societies differ in their bureaucratic relations, as they differ in many areas.
Even such apparently similar societies as those of the 'United States and
Great Britain exhibit sharp differences in bureaucratic interactions. In the
case of police-citizen relations, for example, compared to their British coun-
terparts American police tend to exercise control more informally, and
American citizens tend to have less respect for the law, and to expect less
considerate treatment from police. 3
An important consideration in bureaucratic relations of a technologically

180
 The Broader Context of Bureaucratic Relations
advanced society is the extent and persistence of different subcultures and
classes, particularly as expressed in clients' preparation and readiness for the
impersonalism, hierarchy, and institutionalization of bureaucracy. 4 In this
connection bureaucrat-client relations in the United States may be said to
reinforce and be structured by the American system of persistent and cumu-
lative inequalities experienced by subordinate groups, particularly as this
system consigns people to poverty and low expectations of mobility, monop-
olizes the service fun.ction, and grudgingly provides for public services.
In stating this one must go beyond observing that the character of client
treatment at the hands of street-level bureaucrats reflects and reinforces
social class and racial divisions. While some may be tempted to view the
character of U.S. social services as an expression of racism and of attitudes
toward the individual in mass sodety, we cannot conscientiously stop there.
This view does not explain the development of such an elaborate service and
control apparatus. Nor does it account for the opportunities clients have to
redress grievances, find support within the service network, and resist dehu-
manized services. Nor does it account for or comprehend the wide range of
forms and structures affecting clients in street-level bureaucracies through-
out the country.
Moreover, asserting that street-level bureaucracies reinforce social cleav-
ages does not begin to account for the content ofbureaucratic behavior. Cop-
ing behaviors and adaptive attitudes may be endemic in organizational life.
But this says nothing about the nature of the coping behaviors, or the orien-
tation of adaptive attitudes. Street-level bureaucrats do not fully invent re-
sponses to work stresses but instead at least partially reflect the culture in
which their agencies are embedded. In other words, responses to work
stresses arise out of the work situation, but their content or direction are col-
ored by prevailing cultural assumptions.
In what ways do street-level bureaucracies reflect and perpetuate the val-
ues of the larger society? There are at least two respects in which the struc-
ture of relationships between workers and clients appears to be derived from
the particular character of American society.
First, street-level bureaucracies are affected by the prevailing orientations
toward the poor in the United States. These orientations include the deep
conviction that poor people at some level are responsible for the conditions
in which they find themselves, and that receiving benefits labeled "for the
poor" is shameful. These convictions are epitomized in the observation that
public programs for poor people are almost always treated in the pres:; as
costs to society, not benefits.
These attitudes toward social services for the poor amount to a general
THE FUTURE

stigmatization of poor people. Stigma leads to a general reluctance to join the

deviant group in the society on the one hand, and on the other hand pro-
vides subtle justification for patterns of practice that result in inadequate ser-
vice provision. Prevailing attitudes toward the poor permit rationalization of
patterns that result in client neglect, which would be more difficult to ratio-
nalize if clients were middle class and generally respected. The same may be
observed in agencies of training and control. Some lower courts and public
schools, for example, develop community reputations for dealing mostly
with low-income clients, and they develop patterns of practice that process
people less respectfully than similar institutions with middle-class clienteles.
Intersecting with attitudes toward the stigmatized poor are attitudes
prevalent in the larger society regarding clients' racial or ethnic back-
grounds. Racism also affects the extent to which public employees regard
clients as worthy, and it affects the extent to which patterns of practice
evolve that distinguish among clients in terms of their racial backgrounds.
Second, the politics of the larger society affect street-level bureaucracies
and their clients in the dynamic relationship between the requirements of
providing services and their perceived costs. Governmental initiatives for
programs of social service and control expand or contract, grow more quickly
or more slowly, in part depending upon the relative degree of concern over
crisis or control. In periods of social turmoil or widely perceived crisis (the
depression of the 1930s, the ghetto revolts of the 196os, the "Sputnik" crisis
in education in the 1950s) service benefits and/or funds for training and con-
trol functions increase. In periods of relative quiescence pressures are ex-
erted to return the balance to a ratio of benefits to costs more favorable to
costs. Social analysts may disagree on the precise dynamics of the dialectics
of expansion and contraction of governmental social service and control poli-
cies. But there should be little doubt that public bureaucracies that normally
process clients vacillate in their generosity toward client treatment. Street-
level bureaucracies are alternately able to treat clients with greater degrees
of latitude and forced to restrict options and more narrowly designate
benefits. 5
In the current period street-level bureaucrats are under pressure to de-
velop more restrictive patterns of practice. They are under pressure to in-
crease case loads and to be more formally accountable, and they are gener-
ally asked to expand or maintain coverage in the face of static or declining
budgets. Talking to any social worker or public-school teacher, one discovers
that over time the formal reporting requirements of the job have increased
iu response to managers' efforts to secure greater control. The opposite may
be observed when programs are attempting to attract clients; then s "et-
 The Broader Context of Bureaucratic Relations
level bureaucrats are encouraged to open more widely the categories of pos-
sible assistance and to render aid in more open and responsive ways. 6
It seems apparent that American street-level bureaucracies must be un-
derstood as organizational embodiments of contradictory tendencies in
American society as a whole. The welfare state calls for and requires .social
programs to ameliorate the neglect and insecurity of the economic system, to
prepare people for roles in the economy, or to manage their deviation from
expectations of appropriate behavior. 7 In the ideology of the welfare state
humanitarian impulses are coincident with the requirements of system
maintenance.
This, of course, begins to explain how people with humanitarian impulses
can work for impersonal, paternalistic, or repressive public service agencies.
Most people never question that the requirements of the state are congruent
with the needs and interests oflarge numbers of people. Thus, teachers with
compassion for children work in brutalizing schools and picture themselves
as victimized by the same system that victimizes their pupils. Social workers
with compassion for poor people participate in assigning inadequate benefit
levels to welfare recipients and wish they could do more. 8
The legitimacy of the political and economic system depends on the ap-
pearance of providing for those who cannot provide for themselves and
responding openly and fairly to citizens' claims. Public service workers ac-
tively translate this requirement into programs. But government policy is
not likely in fact to respond fully to the needs of citizens for at least two
reasons relevant to this discussion. 9
First, there is no agreement as to what those needs are. What it means to
"respond fully to citizen needs" is a socially determined concept albeit de-
fined by a process that gives more weight to policy elites than clients. As
continuing controversies in such areas as health care, welfare, and legal ser-
vices reveal, the demands of citizens are open-ended while program costs
must be kept within certain bounds. Indeed, the definition of those bounda-
ries is the basic issue in social welfare policy making.
Second, there is a powerful imperative to maintain private responsibility
for social needs and to make dependency punishable by welfare, public hos-
pitals, and inner-city schools. Granted that street-level bureaucracies exist
outside of the welfare context, and that limitations on program expenditures
must be encountered at some point. Yet it is not at all clear that the United
States inevitably had to develop relatively low social service and benefit
levels compared to other advanced industrial countries. 10 In an assessment
of income and service provision an independent role should be assigned to a
perceived need among policy-making elites to limit benefit and service pro-
THE FUTURE

visions, allegedly to enhance individual and family self-reliance and to stig-

matize the status of worklessness and poverty.
In short, this is a political system that, whatever its current levels of social
welfare expenditures, must also symbolically project images of adequate and
reasonably comprehensive social welfare programming to taxpayers and
middle-class consumers, while in fact it limits support and assistance. 11 Such
a system develops mechanisms to maintain legitimacy and deflect criticism
that the society does not provide adequately for its citizens. Street-level bu-
reaucrats mediate between citizens and the state in that clients' inability to
obtain benefits or services and inequities of distribution may be understood
by clients as personal malfeasance of street-level bureaucrats or administra-
tive agency disarray.
The mediation of street-level bureaucrats helps structure the nature of
urban conflict in several ways. First, as the deliverers of policy they per-
sonally must build into their work life responses to decisions made at the sys-
tem level. Thus a teacher with a classroom of forty students must develop
ways to cope with a high enrollment. Although the teacher did not decide to
attempt education with this student-teacher ratio, he or she nonetheless
must confront this condition. Parents and children, however, may perceive
teachers' coping responses in their own terms and attribute the quantity and
quality of educational services to teachers' abilities or their own capacities to
utilize teachers' services. They may not look to the systemic decisions that
relegate them to large classes, particularly as these decisions are usually
screened from the view of the typical parent and student.
Street-level bureaucrats' needs to control their work situations force them
to defend themselves and the current arrangements. Agency expectations
and occupational norms preclude the excuse that working conditions prevent
effective efforts on clients' behalf, despite private recognition that this is the
case. This defensiveness separates street-level bureaucrats from their poten-
tial allies in improving working conditions for mutual benefit.
Second, the difficulty of controlling street-level bureaucrats is sufficiently
understood by clients so that issues of service are translated into issues of
controlling workers or holding them accountable for performance. Agencies
may deflect potentially significant demands by directing attention to ques-
tions of individual service provision, without examining the structural basis
for those relationships.
Third, it seems easier (although it actually may be harder) to change work-
ers' behavior or approaches to their jobs than to affect the political system
that structures the jobs. Clients and client groups appear to have a better
chance to reorient (reeducate?) individual street-level bureaucrats in a single
 The Broader Context of Bureaucratic Relations
facility or field station than to affect the general patterns of recruitment,
budget making, and policy establishment that condition the work. These
other aspects of the work structure are determined far away in time and
place from the grievances experienced by clients. How much easier it is to
meet with workers and discuss changes in routine, procedure, and attitudes
than to organize to work with others to change the policy system.
Thus street-level bureaucrats' mediating role functions beyond providing
face-to-face contact and facilitating potentially flexible and appropriate re-
sponses among lower-level workers. It also helps define problems citizens
have in dealing with public service agencies as conflicts associated with
lower-level workers. 12
In addition to absorbing conflict in their buffer roles, street-level bureau-
crats in other ways help shape the general attitudes of clients and workers
towards public services. 13 First, the structure of street-level practice has an
impact on clients' conceptions of their capacity. Street-level bureaucrats'
control of clients confirms for clients their dependency and subordination. In
tum, this control more generally affects clients' self-respect and self-expecta-
tions 14 (see chapter s).
Second, street-level bureaucracies discourage employees who seek to
work as advocates. The process begins when the agencies provide an avenue
for people with altruistic orientations to enter the work force. Although job
security is a significant attraction, another important element in seeking
street-level work is the opportunity to help people. Teachers, social work-
ers, legal aid lawyers, and police officers all enter the work force at least in
part with a desire to make a contribution to individuals or to the community.
In some fields public agencies have a monopoly on jobs available to people in
certain professions. Social workers may be able to seek employment in pri-
vate as well as public agencies, but young adults aspiring to become police
officers or teachers largely have to seek public employment if they want to
work in these areas.
Once attracted to these occupations, however, the dynamics of street-
level bureaucracies combine to persuade workers that they are destined to
be ineffective in their chosen fields, that clients may not substantially benefit
from their efforts, or that conditions of successful intervention are not likely
to be available. These conclusions are all the more persuasive because they
appear to be substantially true, at least in the short run. It is difficult to aid
clients in ways consistent with idealized conceptions of assistance within
street-level bureaucracies as they are currently structured, particularly
when the least experienced workers are thrown into the most difficult work
environments.
THE FUTURE

Thus, generations of thoughtful and potentially self-sacrificing people are

disarmed in their social purpose. They come to believe that it is impossible
to find conditions conducive to good practice, and that public agencies can-
not be otherwise structured. Their choices appear to be to leave public
employment for other work or to resign themselves to routine processing of
clients while instructing the next generation of idealists that there is little
sense in hoping for change or in rendering human services.
Similarly, the practice of street-level bureaucrats leads to the self-fulfilling
prophecy that relations with clients cannot change. The actions of street-
level bureaucrats confirm for clients that they will continue to be treated as
they have always been treated. This perpetuates the cycle of the irrelevance
of professional help and reinforces tendencies toward despair and inaction.
This is the most painful part of the estrangement of workers from their origi-
nal purpose. These orientations reinforce the tendencies originating in the
culture toward enhancement of private interests and the abandonment of
social purpose.
If the potential for a more humane practice were entirely negative, of
course, these considerations would only be lessons in realism. However, the
relevant experiences are not monolithic, and the impressions of hope-
lessness are as much socially constructed as they are rooted in fact. The po-
tential for growth and change in street-level bureaucracy is dependent on
both identifying the critical problems and recognizing that patterns of prac-
tice may be reconstructed as well as reproduced. With this in mind it may be
useful to mention several other factors that contribute to the difficulty of
reconstructing street-level bureaucracy.
First there is the problem posed by the apparent conflict between con-
tending interests. Street-level bureaucrats have different interests from
clients, organizational superiors, and the public to which they are nominally
subordinate. The conflictual nature of these relationships is particularly ap-
parent in the current fiscal crisis, which appears to call for taking one side
against another. Conflict between contending interests may be inevitable
(by definition), and in any event it may be socially productive, 15 but it is dif-
ficu'lt to think about deliberately changing major institutions when it appears
that the interests of one party conflict so fundamentally with the interests of
other powerful groups. The more that contending interests appear to be fun-
damentally in conflict with others, the more hopeless social change appears to
be for interests that are relatively weak.
If, for example, parents believe that changes they favor conflict with the
interests of teachers, and that teachers are powerfully situated in the school
system, they will be discouraged from attempting to make changes. Simi-

186
 The Broader Context of Bureaucratic Relations
larly, teachers who believe they are not supported by parent groups or by
school authorities (or by the public-at-large) will be unlikely to seek changes
in working conditions that go beyond those expected of normal labor-
management demands. In short, the more different groups perceive them-
selves to be in conflict with others, the more narrowly they must choose
among objectives and the less they can devote themselves to long-run goals.
A second difficulty is that the patterns of practice that develop in this work
are rooted in the fundamental coping requirements of the job. These are not
easily abandoned or changed because they are experienced by workers and
outside observers as virtual job requirements. People do not readily give up
survival mechanisms. This is one of the reasons it is easier to change articu-
lated policy from the top than to change practice from below. Policy articu-
lated from the top is not rooted in defense mechanisms developed to cope
with the job, while the policy that emerges from practice is rooted in
survival.
Curiously, a third reason it is difficult to think about changing street-level
bureaucracies is related to the appearance of flexibility and innovation.
Street-level bureaucracies are permeated with turmoil rationalized as
change-related. The social service industry of managers, management con-
sultants, public administrators, foundation officials, and academics whose
business is to tinker with social service improvement insures that public per-
ception of street-level bureaucracies is one of constant alterations in the
structure of service delivery. The volume of pilot programs, demonstration
projects, and innovations in management and personnel practices presents
the image of frequent reconstitution of public agencies. This profligacy of
reform confuses a public that cannot possibly assess the programs under-
taken in the name of reform and rarely experiences results of such reforms
directly. The pluralism of the social service industry tends to credit virtually
any kind of change so long as it has its origins in a legitimate source (which
rarely includes clients). Thus the currency of reform is debased. With so
many reform proposals no one change seems better than any other change.
Finally, thinking about significant changes in street-level practice implies
a commitment to altering or improving relations between individual workers
and clients. Yet we are profoundly shy and inexperienced in talking about
relations between and among people. We know much more about deploying
resources than about affecting working relations. It is typical for community
meetings to address issues of recruitment, procedures, incentive structures,
chains of command, and so on, instead of confronting the problems that actu-
ally brought people together in the first place-incompetent or insensitive
teachers, police officers, or social workers. It is easier to avoid these prob-
THE FUTURE

lems, the heart of community relations, and defer them as professional mat-
ters better left for professionals to handle. In any event it is difficult to
measure the quality of relationships; better to stick to dimensions of the
work more subject to administrative manipulation.
In sum, if bureaucracies mirror the society in which they develop it is dif-
ficult to change bureaucratic forms fundamentally without larger changes
taking place. While superficial changes may frequently develop in bureau-
cratic organizations-for example, changes in the relative degree of adminis-
trative centralization-in a sense there remains a deeper structure of organi-
zational relationships that are not likely to yield easily to administrative
rationalizations. This provides an additional reason for insisting that policy
implementation in street-level bureaucracy must be studied at the work
place rather than tracing policy through the bureaucratic and interorganiza-
tional systems. Not only do street-level bureaucrats exercise discretion to
such an extent that they are not easily affected by policy articulation from
above (as I have previously argued). Also, the character of worker-client and
worker-supervisor relations, no matter what the articulated organizational
policy, is likely to continue to reflect the dominant bureaucratic relations of
the society, no matter what the administrative guidelines provide.

Contradictory Tendencies in Street-Level Bureaucratic

Relations

By their nature political systems are relatively impervious to change. To

identify a set of social relations as a political system is to draw attention to the
relative stability of the patterns of interaction that make it up. While it is
useful to specify as clearly as possible the forces supporting the status quo, it
is hardly surprising to discover that the prevailing structure of relations is
deeply rooted.
However, just as the larger society establishes the environment in which
bureaucracies operate and significantly affects bureaucratic relations, it also
gives rise to contradictory tendencies that provide opportunities to challenge
prevailing relationships. Going against the grain, these tendencies are more
difficult to identify and more hypothetical, since their contributions are
rarely dramatic. Nonetheless, consideration of the prospects for change in
any political system requires specification of tendencies that potentially give
rise or support to reformation.
Of the attributes that support change in street-level bureaucracies at least

188
 The Broader Context of Bureaucratic Relations
five should be noted. First, public programs of entitlement and control pro-
vide at least the potential for mobilizing clients and sympathetic publics
toward greater accountability in implementation and administration. 16 One
of the contradictions of the American service sector is that programs con-
tributing significantly to the management and control of populations are sup-
posed to be responsive to public preferences. This, of course, does not mean
that democratic theory necessarily calls for the citizen-subjects of public
agencies to control agencies. Agencies of service and control are theoreti-
cally accountable to a larger society, of which the client population is only
one, and an often relatively powerless, constituency. Still, while public pro-
grams have many constituencies, they more or less present the potential for
formal responsiveness to client preferences. The requirements of account-
ability in theory contribute to the actual capacity of client populations to or-
ganize to change public agencies as they strive to close the gap between ac-
countability in theory and practice.
This consideration is likely to be important in proportion to the population
covered by service. One possibility is that public agencies will simply dif-
ferentiate among high- and low-status clients. But another is that service will
improve for all if high-status clients are included in the population mix. As
public-health care delivery becomes more and more generalized and less the
concern oflow-income populations, it is more likely that clients will be able
to have an impact on service quality. Likewise, parents of children in an in-
tegrated school can have a greater impact on the quality of service than can
parents of a segregated school whose needs can be more easily isolated.
The importance of this consideration increases as the public service sector
in the society grows. As people are thrown more and more into the public
sector for relief or entitlements they will be more able to make claims based
on the discrepancies between programs that would provide for them ade-
quately and programs as they actually exist. Government becomes more vul-
nerable to challenge as it assumes more responsibilities. This is why the ser-
vice state, increasingly penetrating aspects of private life, contains elements
that contradict the control it presumably exerts.
Second, professional norms of behavior toward clients provide a measure
of resistance to bureaucratization. Street-level bureauc1 ats' claims of profes-
sional status imply a commitment that clients' interests will guide them in
providing service. The implicit bargain between the professions and society
is that in exchange for self-regulation they will act in clients' interest without
regard for personal gain and without compromising their advocacy. In a
word, according to professional norms professionals' interest as expressed on
behalf of clients must be unalloyed.

189
 THE FUTURE

This is not to say that street-level bureaucrats do not also confront organi-
zational demands. On the contrary, the essence of their dilemma is that they
are partly professional and partly bureaucratic. However, the potential for
appealing to the professional dimension of these work roles means that there
is an irreducible minimum consideration of the importance of respecting
clients' individuality and acting accordingly.
Even in the roles of police officers and judges, where the concepts of
client advocacy and representation might appear remote, we can see norms
of appropriate application of sanctions that in these occupations are the
primary ingredients of resistance to bureaucratic treatment of offenders. Al-
though it is tempting to see teachers, social workers, and other low-level
workers as constantly under attack for routinizing social services, one can
tum this observation around by calling attention to their resistance to bu-
reaucratization (to the extent that they have been able to resist), largely
stemming from their recognition of the importance of this aspect of profes-
sional identity. (The role of professionals in reform of street-level bureau-
cracy is discussed at greater length in the next chapter.)
Third, street-level bureaucrats by definition interact constantly with
clients. This provides the salutory condition that workers must continually
attend to the people they are supposed to serve and their problems. How-
ever elaborate the defense mechanisms developed to shield themselves from
the enormity of clients' needs, street-level bureaucrats at some level retain a
sense that the people with whom they come in contact are not sufficiently
served by the agencies designated to do so. Thus, one might speculate that
street-level bureaucrats more than other organizational workers are able to
retain a concept of the notion of need in relation to what is actually being
provided. This residual awareness may provide a resource that can be
tapped.
Fourth, lower-level workers maintain a degree of control over their work
environment. Individually street-level bureaucrats exercise discretion to
control the work situation. Collectively many street-level bureaucrats are
able to have a significant say in the rules under which they are employed.
Particularly at the individual level this discretion is not likely to be signifi-
cantly eroded so long as street-level bureaucrats' jobs require them to make
discretionary judgments that cannot be entirely programmed.
Finally, there is a distinct but neglected precedent for organized public
employees championing the needs of clients. Teachers have included limita-
tions on classroom size as an objective to be sought through collective
bargaining. They have sought this objective not only to improve working
conditions but also to create the environment in which they could function

I go
 The Broader Context of Bureaucratic Relations
optimally as teachers. Likewise, social workers have struck on behalf of
improved benefit levels for clients. 17
The cynic may wish to point out the strategic advantage to public workers
of couching bargaining objectives in altruistic terms (although managers are
equally guilty, insisting that they act on behalf of taxpayers and the economic
well-being of the community). Still, cynical or not, such alliances, the stuff of
politics, may be exploited by client groups, particularly when, in collective
bargaining in the fiscal crisis, wage gains are subordinated to improvements
in working conditions.
The impulse to provide fully, openly, and responsively for citizens' service
needs exists alongside the need to restrict, control, and rationalize service
inadequacies or limitations. This is the central contradiction of social ser-
vices. It is more than simply a tension between costs and benefits. It is criti-
cal to reassure mass publics that their elemental needs will be taken care of if
they are not met privately and to rationalize service inadequacies by deflect-
ing responsibility away from government.
Through street-level bureaucracies the society organizes the control, re-
striction, and maintenance of relatively powerless groups. Antagonism is
directed toward the agents of social services and control and away from the
political forces that ultimately account for the distribution of social and mate-
rial values. Thus the American system of service delivery and control is
shaped by the aspirations of the population and by the requirements of the
larger political and social system. In this sense the United States, no less
than other political systems, lends public bureaucracy its particular
character.
 CHAPTER 13

Support for Human Services:

Notes for Reform and
Reconstruction

I have argued that the determinants of street-level practice are deeply

rooted in the structure of the work. Further, I have pointed out that street-
level bureaucracies do not stand alone, but they reflect the character of
prevailing organizational relations in the society as a whole. In turn, as a
primary instrument of contact between government and citizens, street-
level bureaucracies reinforce the relationships between citizens-both
clients and workers-and the state. These observations contribute to our un-
derstanding of the stability of the institutions and their unlikely respon-
siveness to significant reform activities.
Nonetheless, it is important to address the potential for significant reform,
however remote. To say that institutions are stable does not mean that they
are inert, or that the possibility for movement is unavailable. Indeed, street-
level bureaucracies continually confront proposals for change. Seeking ef-
ficiency, equity with flexibility, and appropriateness of intervention, from
different per$ectives public officials, client-oriented interest groups, orga-
nized public employees, and policy analysts perpetually engage in activities
to reform the public services. A theory of street-level bureaucracy should
help clarify the stakes in and potential for reform perspectives.
At any given level of public support we seek at least three values from ser-
vice bureaucracies. We seek services or benefits appropriate to our situation
 Support for Human Services
or needs, equity tempered by flexibility in the distribution of public bene-
fits, and respect as citizens receiving our due from government. Many of the
criticisms of street-level bureaucracies focus on the extent to which people
fail to receive appropriate, equitable, or respectful encounters. Taking these
criticisms as points of departure, three major lines of analysis are discussed
below.
1. Encouraging client autonomy and influence over policy.
z. Improving current street-level practice.
3 Helping street-level bureaucrats become more effective proponents of
change.
There is a necessary and inevitable tension between the desire to have an
impact in the short run, and the recognition that problems are not reducible
to short-term incremental manipulations. Furthermore, significant changes
in street-level bureaucracy are likely to be realized only in the context of
social changes that support the relationships that must be forged. Short of
such changes, these lines of analysis simply become points of dispute in an
ongoing struggle over the relationship of citizens to the state.

Directions for Greater Client Autonomy

Proposals for greater client autonomy generally suffer from the fact that
clients tend to remain relatively powerless. Clients accorded greater collec-
tive influence may not possess the bureaucratic skills necessary to operate in
the policy arena, or they may inherit control over programs or facilities so
bankrupt that they defy significant management improvements. Tenant
management of underfunded and poorly maintained public housing, for ex-
ample, may fail to improve service more for financial and structural reasons
than for reasons of client capability. In this sense giving clients control over
public facilities may contribute to social control, as tenants contest with
other tenants over scarce jobs and project resources.
Nonetheless, proposals to increase client autonomy must be vigorously
studied for their potential contribution to changing street-level rela-
tionships. One approach to change in street-level bureaucracy is to eliminate
public workers as buffers between government and citizens. A class of pro-
posals utilizing such an approach is represented by plans to issue service
vouchers to citizens. By providing clients with claims on public or private
service agencies, sponsors of voucher proposals hope that agencies would be
more responsive to client preferences in order to attract their patronage.

1 93
THE FUTURE

Voucher proposals have gained sup\'ort because they promise to introduce

consumer sovereignty into the production of social services.
After considerable experimentation with educational vouchers the record
suggests that it is extremely difficult to establish the conditions under which
clients of educational services are fully informed about a wide variety of edu-
cational options. It is not so much the theory that is damaged by these expe-
riences as it is the hope of creating the rudiments of competition on which
the theory actively depends. 1
On their face, voucher proposals are attractive because they evoke the
model of a competitive market that develops products in response to con-
sumer demand. Unfortunately, market models in service provision will not
solve any problems so long as service providers monopolize the scarcely
supplied skills of semi-professionals, dictate the conditions under which ser-
vices will be supplied, or are allowed to limit information available to the
service consumer. Moreover, even in theory market models can only be as
appropriate as clients can be expected to have an opinion about service qual-
ity. This creates confusion in areas such as health and even education, where
clients cannot always assess the appropriateness of service.
There is surely something attractive about the idea of providing people
with money to purchase services on the open market. But so long as profes-
sionals control access to services prices will be bid up, and bureaucracies will
be created to control eligibility and costs or insure minimum standards.
Moreover, variation in location is a critical aspect of access to service, repre-
senting a substantial hidden cost to some service consumers. Segregation of
clients by classes would likely continue to take place unless ways were found
to overcome the costs of seeking services on a geographically diffuse basis.
Another set of reform proposals calls for eliminating mediating public
workers from service contexts which, properly supported, might be handled
by citizens with little or minimum assistance. Thus legal reformers encour-
age experimentation with community dispute resolution and simplified legal
procedures in order to make the law accessible to citizens without lawyer
mediation. Community dispute resolution mechanisms presumably free the
courts and other legal institutions from cases that might be processed by
other means. Similarly, home care programs in the health field represent ef-
forts to permit people to avoid institutionalization and make maximum utili-
zation of service personnel. On an experimental basis many school systems
attempt to provide educational experiences through utilization of commu-
nity resources. Citizen patrols have developed to end or supplement com-
munity dependence on the police.
Not all proposals to support indigenous efforts to provide service and elim-

194
 Support for Human Services
inate mediating public employees will accomplish these objectives. They
may spawn quasi-public agencies that have the potential for replicating the
difficulties of the agencies they replace. They may develop entitlement, reg-
ulatory, or service bureaucracies that perpetuate bureaucratic experiences.
Home care, for example, frees people to stay out of the hospital under cer-
tain circumstances, but it still requires a bureaucracy to certify eligibility, to
promulgate and monitor standards for service providers, and to see that ser-
vice personnel are hired to provide home care. In general, questions of
supply and maintenance of standards remain in all service areas so long as
the government retains ultimate responsibility.
These alternative perspectives on service provision suggest opportunities
to define the relationship of providers and clients differently. However, they
do not fundamentally reorganize the need for service in many instances and
do not offer guidance where street-level bureaucrats remain in a controlling
relationship. A sharp need continues to provide a better balance of power
between street-level workers and clients. A better balance would be
achieved if the following developments were encouraged.
Wherever possible, opportunities should be seized to demystify street-
level bureaucracies and the practices in which they engage. Workers should
be taught how to communicate with their publics in plain language, and
clients should demand explanations they can understand. Client advocates
should be sponsored and trained to guide clients through the bureaucracy,
to obtain answers they are otherwise unable to get, and to represent clients
to workers where they would otherwise be intimidated. Guides to clients'
rights and maps of bureaucratic systems should be developed; more impor-
tant, street-level bureaucracies should simplify procedures to make service
systems more manageable without expert intervention.
Simple practices should be developed to make street-level bureaucracies
more accountable to clients. Requiring workers to provide summaries of the
transactions clients experienced but may not have fully comprehended
would be a significant step forward in some places. Routine reviews to deter-
mine whether clients were receiving all benefits to which they ~ere entitled
would place the burden of programming for clients on public employees.
Such details would modestly contribute to the development of more recipro-
cal relationships.
As a matter of public policy we should welcome investigations by public-
interest law firms, legal services offices, government agencies, and others
challenging prevailing practices where those practices entail responsible
allegations of inhumane service or systematically neglected clients' rights.
We should recognize that the discretion of street-level workers is uncer-

195
THE FUTURE

tainly monitored at best and that governments that create these bureau-
cracies may properly oversee their direction by encouraging client as well as
bureaucratic scrutiny.
The struggle of clients to organize and obtain some control over service
provision should be respected and encouraged. Client involvement in gover-
nance of service agencies will help to insure that clients contribute to the
way street-level bureaucrats define their roles. Service provision should be
decentralized to a significant extent, so that the advantages of orienting prac-
tice toward local initiatives can be realized.
During the 1g6os some communities experimented with client partici-
pation in governance of schools, neighborhood health centers, public hous-
ing, and other public services. From these experiences we should know bet-
ter than to encourage citizen control without examining the conditions of
transfer and the degree of control. More often than not, the experiments of
the 1g6os inappropriately discredited citizen participation by providing con-
trol over programs lacking financial viability or by narrowly circumscribing
the scope or powers of client or citizen boards. While avoiding the problems
of cooptation of community activists in financially unhealthy public en-
terprises, or of tokenistic participation, client control over service bureau-
cracies remains potentially critical in making the bureaucracies more respon-
sive to clients.
Issues of client control over service bureaucracies are not separable from
considerations of large-scale social changes or changes in the organization of
public services. It is surely simpler to contemplate neighborhood represen-
tative councils governing outreach health centers, legal services programs,
and local schools than it is to consider citizen participation in the governance
of centralized health facilities, downtown legal services complexes, and con-
solidated schools. Yet, if client control represents a way of making clients
more central to street-level bureaucrats' tasks and promoting further
changes by fundamentally challenging the bases of service provision, then
extending citizen control over complex systems and facilities should be
encouraged.

Directions for Current Practice

TAKING DISCRETION OUT OF BUREAUCRACY

Managing discretion is at the heart of the problem of street-level bureau-
cracy. For the most part, society is not willing fully to circumscribe street-
level discretion. However, there may be some contexts in which it is desira-

196
 Support for Human Services
ble to circumscribe it. It is hardly obvious that every discretionary role
played by street-level bureaucrats should continue to exist. Where workers'
discretion leads to unfair and unequal treatment of clients, with no compen-
sating benefits, it should be desirable to reform systems by removing this
unredeemed source of unfairness. If, for example, a teacher's right to strike
pupils is judged to be undesirable in all cases, it is obviously appropriate to
prohibit flatly this ?Ption.
Some situations may arise, or may develop from previous practices, in
which the judgment must be made that intervention by street-level bureau-
crats is harmful or wasteful. However, the judgment that street-level bu-
reaucrats' discretion is inappropriate is not necessarily easily made.
The decision to separate social service functions from income-support de-
termination functions in public welfare is an example of removing consider-
able discretion from street-level bureaucrats. The stated intention was to
limit discretion of case workers in support determination and free them to
provide services without the burden of client appeals or tactics designed to
obtain more benefits. (Perhaps the primary appeal of this "reform" was that
it also promised to reduce the number of welfare personnel.) This change in
policy probably deserves support because the capacity of social service work-
ers to provide meaningful assistance to clients had become so circumscribed
that their interventions had largely lost whatever beneficial potential may
have existed in the previous definition of social worker roles. Moreover, the
paternalistic and degrading assumptions of the welfare system virtually
guaranteed that social services would be provided in a nonreciprocal con-
text. In this case the discretionary opportunities of case workers had already
become so limited by the social context in which aid was provided that it was
probably beneficial to further routinize the aid process. 2
Still, some doubts linger. Questions remain concerning the meaning of
this "reform." To what extent was this policy adopted in order to reduce the
power of social workers, diminishing their role and also their numbers? To
what extent was the transformation of support determination motivated by a
desire to reduce support levels by eliminating client-oriented social workers
from the process in favor of cost-conscious accountants? To what extent did
this reform indeed eliminate discretion, or did it simply transfer discre-
tionary powers to a new set of employees? Finally, by eliminating the social
service provider at intake, to what extent did this reform represent an im-
plicit decision to reduce service levels by restricting social workers' opportu-
nities to pick up relevant problems at intake, when people are often most
receptive to assistance?
Some of these questions arise generally in considering systems to rou-

197
THE FUTURE

tinize discretionary judgment. Hospitals attempt to develop elaborate pr~

tocols to help nurses determine medical priorities in emergency rooms. This
is done in the name of optimizing the use of available resources. But the as-
signing of priority categories also restricts the observations that can be made
about a patient. Responding to the most salient symptoms may mean neglect
of the whole patient or overlooking other conditions requiring diagnosis.
Similar problems with categorization may be observed in legal services.
To serve more clien'ts legal services offices often train receptionists to dis-
cover quickly the reason for a potential client's inquiry, and then refer the
caller to a trained paraprofessional. Systems are elaborated to help the re-
ceptionist do this reliably. For this system to work well the receptionist
should precisely follow the guidelines for referral.
Two problems arise in such a system, however. First, it is extremely dif-
ficult to circumscribe all possibilities. Thus the receptionist needs a back-up
to handle calls that are not easily categorized. Second, and much more dif-
ficult to resolve, the presenting problem may be only the client's most
urgently felt concern. The client may have more pressing and perhaps easily
resolvable legal problems, but he or she is not aware of them or aware that
assistance may be availabie. The system of early categorization, dependent
on nondiscretionary slotting by the receptionist, produces efficient use of
paraprofessional resources at the possible expense of serving the clients
holistically.
On balance the judgment may be made that the costs of categorization are
more than compensated for by the benefits of standardization. But it is a
judgment that must be made with recognition of the liabilities involved. The
logical extension of this way of thinking is to eliminate people from these
judgments entirely. It is technically possible to program computers to make
categorical determinations and even to advise people on, say, errors they
may have made on their applications. This way of thinking similarly mo-
tivates adherents of computer-aided teaching machines, which may be pro-
grammed to provide the closest thing to a warm hug that machines are capa-
ble of (talking to students in a friendly tone, asking them personal questions,
providing verbal encouragement). We can make teaching machines and
diagnostic machines. At issue is the importance of seeking aid through other
people. It is likely that when this technology is highly developed the rich
will choose to pay for teachers and doctors, while the poor will be taught by
machines and will obtain medical advice from a computer terminal.
 Support for Human Services
ENHANCING CAPACITY OF STREET-LEVEL BUREAUCRATS
Embedded within the critique of street-level bureaucracy appear to be
piecemeal formulas for reform. If discretion were constricted street-level bu-
reaucrats would have less need for routines and simplifications to deal with
uncertainty. If goals were clearer, workers could direct their energies with
less ambivalence. If appropriate performance measures were available,
street-level bureaucrats could be made more accountable for their behavior.
Within limits these observations are probably correct and sometimes may
form the basis for action. However, ultimately they are likely to be quite lim-
ited for several reasons.
First, conditions of street-level bureacuracy comprise a syndrome. Except
for those instances in which client-worker interactions can be eliminated,
the conditions affecting the work context occur together and cannot easily be
rationalized or simplified. Goals are ambiguous, performance measures are
difficult to obtain, and discretion is required by virtue of the need for human
interaction. It is unlikely that the apparent looseness of the bureaucratic con-
text can be tightened up. If the organization could be tightened up, it is
likely that it would have been tightened up before. To say that human in-
teraction is required in service delivery is to suggest that judgments must be
made about potentially ambiguous situations. Reciprocally, to say that the
conditions of street-level bureaucracy exist is to say that the situation
requires human judgment. It is quite unlikely that a part of the street-level
bureaucracy syndrome can be transformed without a change in the basic as-
sumptions underlying the service policy.
Second, even if it were possible to clarify lower-level workers' decision-
making contexts, it is uncertain whether improvement in one aspect of the
syndrome would alleviate problems arising from the whole. This would be
different if the problems of street-level bureaucracy were additive so that
every diminution of an aspect of the problem would result in a corre-
sponding benefit.
The desirability of increasing or improving the resources of street-level
bureaucrats also seems implicit in the critique. However, as suggested in
chapter 3, we should be extremely skeptical of proposals for additional re-
sources as a solution to problems of street-level bureaucracy. More re-
sources are likely to reproduce problems of service quality at a higher-
volume level. More clients may be served-a valuable objective-but this
may be an achievement comparable to accommodating more cars going
nowhere during rush hour on the Long Island Expressway. Furthermore,
if for some reason agencies are able to hold the line on volume and apply
increases to current client loads, additional resources may not enable street-

199
THE FUTURE

level bureaucracies to cross over a hypothetical threshold level whereby

such resources would begin to have an impact. A reduction in class size from
30 to 25, for example, would require a tremendous commitment of resources
for a large city school system, yet reductions might only begin to be effective
when class size reaches 15 students. 3
On the other hand, maintaining current resource levels may be an impor-
tant objective simply to prevent further erosion of client-worker relations.
To this end the Denver Plan appears to be justified. Under this plan children
are assigned numbers based on the presumed difficulty of teaching them.
Teachers are guaranteed through contract that class size will not exceed the
equivalent of teaching 25 normal pupils. 4 In this way teachers are rewarded
rather than punished for having challenging pupils in their classes. Case
loads in other bureaucracies could be meted out in the same way.
Second, additional resources might not improve working conditions but
instead might increase the number of clients served, dilute the services
provided, or be allocated to the incomes of employees. In the public services
it is currently difficult to hold the line for service quality when there are
clients who remain unserved, demands for more services, and organized em-
ployees who feel aggrieved. Moreover, to the extent that additional re-
sources mean additional worker training it is questionable whether this in-
vestment can overcome the impact of working conditions. Worker training is
less important for practice than the nature of the working conditions them-
selves. 5 Without a supportive network of working peer relationships, train-
ing to improve service capacity of workers is likely to wash out under the
pressure of the work context.
The best chances of affecting work performance through job enhancement
come when the system of service delivery supports workers in maintaining
high standards of service quality. It is likely to be helpful if proposals for sup-
porting practice are specifically job related, that is, helpful in solving specific
challenges experienced by workers. Thus on-the-job training is likely to be
more effective than classroom learning experiences because the training is
provided in the context of actual problem-solvin g situations. Generalizations
about police training apply to other street-level bureaucracies as well. In-
struction relating to experiences police officers are likely to encounter is re-
tained; instruction unrelated to direct problem solving is subject to erosion
because it is irrelevant to officers' needs. In the short run, assistance in job
problem solving is likely to be the most effective way of improving street-
level performance.
This suggests some of the limits to direct job training. If street-level bu-
reaucrats are likely to retain only that counsel relating to current job prob-

200
 Support for Human Services
lem solving, then training oriented toward transforming jobs into something
else is likely to be ineffective. 6 Thus one would be skeptical about develop-
ing new kinds of street-level bureaucrats primarily through instructional
efforts.

The Prospects and Problems of Professionalism

These reform orientations have their limits. Most street-level work is not
open to meaningful revision by limiting discretion, removing public em-
ployees from interaction with clients, or modestly altering bureaucratic
structure. In street-level bureaucracies there is an irreducible requirement
that public employees interact with citizens to determine the nature and ex-
tent of public services they should receive and to provide those services
through interactions with them. If this is the case we are challenged to ob-
tain accountability when bureaucratic mechanisms are inadequate and
inappropriate.
The limits on bureaucracy in this area draw attention to the argument that
street-level bureaucrats should be professionals whose relatively altruistic
behavior, high standards, and self-monitoring substitute for what the society
cannot dictate. Who will watch the watchmen? The watchmen will watch
themselves. The argument for professionalization comes down simply to the
realization that control of occupational groups must come from within the in-
dividual members of the group if it cannot be dictated from outside. If street-
level bureaucrats cannot be restricted in everyday functioning, then self-
monitoring must substitute for bureaucratic controls. To this extent, the
advocates of greater professionalism in street-level bureaucracies appear to
have an unassailable point. There is a powerful coincidence in our apparent
need to have street-level bureaucrats monitor their own performance and in
the aggrandizing claims of street-level bureaucrats that they are indeed suf-
ficiently autonomous and self-policing to be granted at least a degree of pro-
fessional status. 7
The professionalization of street-level bureaucracies is commended by
some analysts because standardized formal training in universities, seeking
training to get credentials, and control over occupational entry is already far
advanced in teaching, nursing, social work, and other street-level occupa-
tions. 8 In addition, it appears possible to influence the professionalization of
these occupations. Public policy to influence the direction of profes-

201
THE FUTURE

sionalization is typically directed toward paying higher salaries to make these

occupations more desirable to a more educated class of people, improving
and subsidizing preprofessional training through universities, overseeing
certification through professional boards to insure minimum standards, and
making promotion and advancement dependent, at least superficially, on
meeting professional standards of performance.
There are some attractions to focusing on professional development for
solving service dilemmas. As discussed above, professionals are theoretically
committed to a service orientation. The professions constantly receive re-
cruits who to some degree are initially committed to realizing these ideals.
Not only do professions embody a service ideal in theory, but also some peo-
ple, believing the theory to apply in practice, seek to enter the professions in
order to have a vocation consistent with ideals of service and sacrifice.
The problem with the "professional fix" in solving dilemmas of street-level
accountability lies in the great gap between the service orientation of profes-
sionals in theory and professional service orientations in practice. The lead-
ing professions of medicine and law have not measured up well to their own
standards, which call for providing service to those in need, dictating the
primacy of the client over the needs or preferences of the professional, and
recognizing that, in light of their monopoly over service, professionals need
to develop practices that meet community requirements. On the contrary,
studies of professional practice suggest that doctors, lawyers, and other pro-
fessions tend to seek out higher-status clients at the expense of low-status
clients, to neglect necessary services in favor of exotic or financially reward-
ing specialities, to allow the market for specialists to operate so as to create
extreme inequalities in the distribution of available practitioners, to provide
only meagerly for the professional needs of low-income people, and to re-
spond to poor people in controlling and manipulative ways when they do
serve them. 9
These are broad generalizations to be sure, and they undoubtedly ignore
contradictory evidence to some degree. Still, one can hardly claim familiar-
ity with professional practice or read the literature on the response of the
professions to social problems without recognizing these observations as cen-
tral tendencies. At the very least these observations caution us to be quite
skeptical of proposals to solve problems of street-level bureaucracy through
increased professionalism.
Undoubtedly many people enter the professions seeking primarily the
high status and income that they have been led to believe they would enjoy.
But what about those who originally aspire to work with a degree of service
orientation and are unable to maintain that direction? Whether they drift

202
 Support for Human Services
away from a service ideal or abandon it after protracted struggle with them-
selves or others, the careers of idealistic professional recruits are usually
abandoned to processes that insure their socialization to the dominant pro-
fessional values. There are at least three areas in which the professions as
currently organized help to erode the service orientation.
First, professionals by definition are accountable only to peers. While the
peer orientation protects professionals from the criticism of untrained and
inexperienced outsiders it also insulates them from the criticism of clients
and people who would speak on clients' behalf. Professionals are notoriously
reluctant to criticize each other and at best direct attention only to the most
extreme violations of ethical norms. Informal peer review is normally
avoided and formal peer review focuses on immoralities unrelated to profes-
sional performance or to narrowly defined technical capabilities. Formal
peer review is irrelevant to guiding the normal routines of practice or the di-
rections of the professions.1o
Of greater impact overall are informal peer pressures (as opposed to re-
views) that guide professional development. Here, as in other social groups,
professional novices are initiated into the rituals of practice through encour-
agement to develop amiable relations with peers, to protect fellow members
of the society, and generally to appreciate the problems of other profes-
sionals, even at the possible expense of clients. Thus the doctor's advocacy of
patients' medical needs is tempered by the requirement that he or she ap-
preciate the organizational needs of the hospital. Thus the lawyer's advocacy
of clients' rights is tempered by his or her need to appreciate the norms of
the court. Untempered peer definition of professional norms thus effectively
erodes the client orientation to which professionals are theoretically
committed.
Second, a major problem with professionalism as a model for street-level
bureaucracy is the tendency of professionals to work in isolation. Effective
professional norms call for mutual deference and shows of respect except
when other professionals grossly violate these norms or threaten to embar-
rass the profession. It is not that well-qualified professionals actually respect
the bumbling fool in their midst. However, it is unlikely that they will take
steps to help the fool improve (or, for that matter, to aid the eager but inex-
perienced novice) except when pressed by matters of self-interest. Looking
at the same problem from the viewpoint of the professional who could use
help, the norms usually inhibit professionals from seeking guidance in solv-
ing problems or providing services to clients, since to ask for help would be
to admit a degree of incapacity. It is instructive that police officers who feel
themselves dependent on their partners for personal safety are extremely in-

203
 THE FUTURE

tolerant of a partner's operational defects. But in street-level bureaucracies

in which co-workers' defects do not so directly impact on others' well-being
(for example, building inspectors), deficiencies may be tolerated for a consid-
erable period of time.
Third, the most powerful agent in professional socialization is the work
setting. Thus it is the extremely rare newcomer who is able to assert un-
popular or unsanctioned values. The education of new recruits as to what is
acceptable, what is appropriate, and what will enhance one's career is an ex-
tremely powerful determinant of future professional behavior.
In summary, although there are both theoretical and practical reasons for
looking to professional development as a way to improve street-level bureau-
crats' performance, the record of the professions suggests that the model
they provide in practice is not necessarily an auspicious one for increasing
responsiveness to clients.

Keeping New Professionals New

During the 1g6os the service occupations and professions appeared to expe-
rience a degree of revitalization as an influx of idealistic recruits joined with
like-minded members of the profession to dedicate themselves to practices
based on different principles from those that were dominant at the time.
These principles included placing their skills in the service of those people
who most needed them (the poor, ethnic, and racial minorities), committing
themselves to respecting client autonomy, turning the power of the profes-
sions to helping achieve greater social and economic justice, and eschewing
personal status enhancement when it conflicted with these principles. These
"new" professionals were radical in the classic sense of seeking a return to
first principles. 11
The surge of idealism during this period differed from other times only in
volume and attention paid by others. In every era people sympathetic to
these principles have entered the professions only to be ground down by the
social structure of their jobs. If there was a difference in kind in the 1g6os it
stemmed from the greater numbers, which provided a critical mass to articu-
late these ideals, and from the social movements that provided the setting in
which these ideals were stimulated and received notice and approval.
In every era, there is a propensity among at least some members of street-
level bureaucracies to work according to the ideal standards of their roles. If

204
 Support for Human Services
we must depend upon a core of street-level workers who will strive to main-
tain integrity in the exercise of discretion, we may well ask what can be done
to support and enlarge this core? Where such a core does not presently exist,
we may ask what can be done to bring it into being? What can be done to
keep the new street-level bureaucrats flexible in their response to clients
and zealous in their commitment to client rights while delivering public pol-
icy? While pursuing these objectives, what can be done to insure that the
new street-level bureaucrat possesses the skills to intervene with clients ef-
fectively? It is helpful to ask these questions in these ways because they
direct attention to building on opportunities that currently exist. They guide
analysis of one possible direction of street-level reform without directly con-
fronting the enormity of attempting to think through fundamental refor~
for the totality of street-level bureaucracies.
The new street-level bureaucrats already exist in public agencies, trying as
best they can to maintain high standards and resist the routinization that
they fear is their fate. The new street-level bureacrats already exist in profes-
sional schools, searching for positions that will permit them to develop ca-
reers consistent with the objectives that motivated them to consider public
service in the first place. The new streevlevel bureaucrats exist in the reser-
voir of young people who would commit themselves to public service if they
had effective service models which they might follow.
What is needed to develop this potential?
Financial support for the human side of street-level bureaucracies is a nec-
essary although insufficient condition. Although additional resources cannot
overcome the patterns of routinization and simplification that are currently
endemic, teachers, legal services lawyers, police officers, and other street-
level bureaucrats will not have the slack to organize themselves for more
responsive interactions with clients unless they have adequate material sup-
port. It is particularly important to reverse the current decline in support for
the human services, since workers' impressions of harassment and resource
inadequacy are probably as important as the fact when it comes time to orga-
nize client processing. Public agencies must provide an atmosphere of delib-
eration or workers will not be able to escape the conviction, rooted in coping
needs, that they must routinize client processing. This is part of the reason
that incremental reductions in case load often fail to-show an effect, since the
feeling of harassment remains when case loads are marginally reduced, say,
from so to 45 cases.
Financial resources are also necessary to provide the incentives necessary
to make possible career commitments. It is not that a new street-level bu-
reaucracy needs to provide the same financial rewards as careers in the

205
THE FUTURE

private sector. However, the substantial material (and purposive) uncertain-

ties ofthis public service work are detrimental to building a cadre of commit-
ted human service workers.
This is less of a problem in the established semi-professions, such as polic-
ing and teaching, where civil service systems and collective bargaining have
resulted in reasonably high salaries. In these areas the competition for places
and current reductions in work forces are larger obstacles to revitalization.
The problem is somewhat different in professions with extremely high sta-
tus. Legal services, for example, has only a very few positions at the top.
Thus young lawyers quickly must determine whether they are going to stay
in this line of work at the considerable risk of a dead-end career or shift to the
professional mainline where financial opportunities are much greater.
It is easier to discuss career ladders and mobility opportunities in periods
of system growth than in periods of contraction. In the current period it is
thus easier to consider molding career opportunities in legal and health ser-
vices than in education, since the work forces in the former are likely to ex-
pand while those in the latter are likely to shrink in response to declining
enrollments.
The new street-level bureaucrat needs rewards for effective performance.
Currently professional associations and public employee unions discourage
performance rewards based upon "soft" evaluative measures. They prefer
so-called merit increases only where the majority of members have easy
access to achieving the rewards. The lockstep of advancement in civil service
systems must begin to yield to assessments of ability based on recognizing
not only the volume of work but also the appropriateness of response and
quality of the interaction with clients. In tum, this will mean seeking out and
crediting citizen reports on street-level workers (with due recognition of po-
tential bias, as is the case of all assessments). It will also mean development
of peer assessments in the provision of services and worker contributions to
determining assessment criteria. Above all it will mean developing ongoing
consultative interactions of workers and supervisors to provide systematic
qualitative evaluations of case handling.
But breaking down the isolation of individual street-level bureaucrats will
be mostly destructive if it is done simply in the name of higher degrees of
scrutiny. The hardest reform of all will be to develop in street-level bureau-
cracies supportive environments in which peer review is joined to peer sup-
port and assistance in the working out of problems of practice. Currently
peer review takes place in the unsystematic culture of lunchroom chatter,
casual observation, and third-hand reputational assessment. Peer instruction

206
 Support for Human Services
takes place in the schooling of rookies in how to survive on the street, with
supervisors, or in the face of client harassment. In other words, peer review
and instruction currently do take place, but in ways that force workers either
to be extremely circumspect or to promote routine processing rather than
responses appropriate to individual clients.
The development of peer support mechanisms can and must be related to
work processes. Street-level bureaucrats need to receive recognition for
good work and to be free to seek help when they encounter work-related dif-
ficulties, without feeling that their reputations are in jeopardy. Perhaps out-
side specialists should systematically review the work of street-level bureau-
crats with their clients. Perhaps agencies can develop this capacity without
such assistance. Whatever the mechanism, those street-level bureaucrats
who continue to aspire to provide appropriate community service will wel-
come the chance to grow in their jobs without being judged and placed at
risk in the process.
A street-level bureaucracy that has developed processes of staff growth
and development will also develop processes for small group decision mak-
ing. Small group units for street-level decision making (for example, grade
levels or departments in schools, subprecinct units in police departments,
neighborhood service offices in legal assistance and health maintenance) are
probably best suited to determine which aspects of social services should be
routinized and which aspects should remain unprogrammed. Routinization
in social life may be inevitable, but it is not inevitable that routines should be
imposed from above or by authorities that do not directly confront clients.
Decentralized units would be far more likely to develop routines consistent
with responsive and efficient client treatment than authorities removed from
the scene, particularly if outside audits are continually able to draw attention
to issues of service quality.
Fundamentally at issue is making the most of the reality that street-level
bureaucrats primarily determine policy implementation, not their superiors.
If the bureaucratic connection between lower-level workers and the reins of
authority are indeed tenuous, as I have argued, perhaps it is better to Aow
with the organizational dynamics of policy delivery in these organizations
rather than resist them by insisting on bureaucratic solutions to problems
defined as worker deviation from preferred performance.
In reality, decentralized units given full responsibility for practice would
have to resist the tendency to drift toward recentralization of routine func-
tioning. The pull would be strong to let higher authorities make critical
decisions, thereby absolving lower-level workers of responsibility. How-
207
THE FUTURE

ever, even creating the opportunity for self-determination of small units pro-
vides a context for considerable learning and the potential for achieving a
more client-oriented practice.
Even if it were possible to restructure street-level bureaucracies in these
ways, what would keep people from drifting into the old patterns? This is a
particularly significant question if we believe that bureaucracy to some degree
reflects as well as reinforces and perpetuates the prevailing social structures.
Three considerations conducive to workers' exercising effective and respon-
sible control over the work situation m,ay help consolidate changes in street-
level bureaucracies in ways that limit the likelihood of retrenchment.
First, the clients of service must become a more potent force in the refer-
ence groups of street-level bureaucrats. Ways must be discovered to make
visible and accessible the behavior of lower-level workers, and clients likely
to be affected by their actions must become more involved in the definition
of good practice. To the extent that peer relations are the primary source of
the expectations and values promoted within the occupational sector, there
will remain a temptation to develop esoteric criteria of practice judgment.
Street-level bureaucrats' performance has not been so good nor our con-
fidence in their work so well established that great harm would come from
creating regular mechanisms to expose street-level work to the scrutiny of
clients. Even if clients are overwhelmed by the trappings and rhetoric of
professionalism or are limited in their understanding of the ramifications of
decision making, exposing the decision-making environment to clients
should anchor street-level bureaucracies more firmly in a client orientation.
Moreover, street-level bureaucrats should undertake to develop techniques
to educate clients toward making better judgments about seeking service
and better assessments of service provision. Studies and observations con-
cluding that clients are overwhelmed by professionals caution us about in-
volving clients in decision making, but they do not reflect experiences in
which client involvement has been systematically nurtured. 12
Client contributions would be enhanced if street-level units accepted re-
sponsibility for group case loads rather than incorporating clients as the case
loads of individual workers. In many street-level bureaucracies, a primary
contribution to workers' isolation and pressure is the fact that workers are in-
dividually positioned to be fully responsible for clients, are unable to seek as-
sistance or advice, and must compete with other workers for advantage so as
to minimize their load. So long as street-level workers are individually re-
sponsible for their sector of client services they are likely to be defensive in
developing cooperative and supportive relations with fellow workers or
clients. Without abandoning the efficiencies of specialization or the account-
 Support for Human Services
ability that individual case loads minimally provide, it is possible to develop
conceptions of group or office case loads that make clients the responsibility
of the staff, not individual workers. 13
A second requirement in sustaining the new street-level bureaucracy is
the zeal and leadership of people committed to the new orientation. Reform
orientations are not self-implementing. They can only survive in a context in
which people are dedicated to public service and receive support from client
groups, fellow workers, and the community. The nurturing of such leaders is
a process that would well begin in the training grounds of universities,
where relatively visionary orientations are sometimes rewarded, and would
continue through a public policy that valued such leaders for their commit-
ment to a client-oriented service. Without the development of such a eward
structure it would be reasonable to conclude that there was no constituency
for these reforms.
Traditionally universities have provided strategic sanctuary for some of
the most important dissenters from contemporary practice, but they have
often been rendered ineffective because their lessons come from the ivy
tower rather than from the streets. This is particularly regretable because
such teachers may inspire young professionals to go into their work without
expe!:iencing the dilemmas of practice or helping to prepare the environ-
ment into which students insert themselves. Ihe new street-level bureau-
cracies would be significantly assisted by policies in which supporters of this
orientation circulate between the teaching of young professionals and the
practice of public service. Some of their teaching ought to be done not in
universities but in the field, where there is opportunity for constant confron-
tation with the realities of practice. A police academy in which students
worked half-time as apprentice officers, or a social work institute in which
students received training from teachers who shared an office practice with
their students, might provide the reality-based environment in which staff
and students would find an appropriate balance between experience and
detachment.
A final aspect of support for maintaining the client orientation of street-
level bureaucracy rests in the development of ongoing processes of suppor-
tive criticism and inquiry. Built into every week of practice should be oppor-
tunities to review individuals' work, share criticisms, and seek a collective
capacity to improve performance. The orientation should be skeptical, for
the objective of such sessions would be to resist where appropriate the early
closure of possibilities that accompanies the inevitable routinization of prac-
tice. Some police departments hold an inquiry every time an officer draws a
weapon. Hospitals ordinarily inquire into circumstances accompanying

209
THE FUTURE

deaths. These inquiries become defensive because the consequences of mal-

feasance are so grave. What is called for is introducing a norm of inquiry into
routine practice, to keep alive the potential of client services and contradict
the neglect that results from ordinary inadvertence. At the same time, the
staff would be more receptive to learning when the potential consequences
were less grave and all staff were in tum the objects of attention.
This general thesis assumes that people who are constantly engaged in
planning for group practice, who have some control over their work pro-
cesses, and who regard clients not as units to be processed but as people,
will discover the rewards of doing a good job gratifying and renewing. The
service sector of the society will be substantially transformed when teachers
find (or rediscover) their rewards in the education of children; police officers
take pride in doing a difficult job strictly on the merits of individual encoun-
ters; and social workers have the resources and the untrammeled commit-
ment to help clients find solutions to their own problems.
There are obvious intellectual hazards in sketching an outline of new
approaches when confronted with the enormity of the present distress in
social services, the ambivalence toward improving social services born of
personal encounters, and the bureaucracies' apparent resistance to-respon-
sive reforms. I will consider it a measure of accomplishment if these notes
toward a new era for street-level bureaucracies are taken as not so un-
grounded in reality that they seem drawn from thin air and not so remote
from human potential that they seem farfetched.
In recognition of the magnitude of the task I have framed this discussion in
terms of building on the commitments of new professionals. But the recon-
struction of street-level bureaucracies is unlikely to take place in the absence
of a broad movement for social and economic justice. Precisely because
reform of mass client processing involves more equitable distribution of ser-
vices as public goods, valuing more highly the status of individuals in soci-
ety, and challenging the control and mystification of public services, it is dif-
ficult to achieve and requires general political support. If street-level
bureaucracies indeed play critical roles in the political structure, isolated
reform efforts cannot plausibly be expected to bear the full weight of social
change.
Current reform interests are fragmented among the three parties of the
buffer relationship. They are divided among administrators who seek to
improve efficiency and effectiveness through management tools; unions that
seek improved working conditions but are constrained to protect job bene-
fits; and clients and client interests that seek service improvements but lack
legitimacy in policy arenas. The quality of street-level practice will change

210
 Support for Human Services
only when an effective coalition develops that harnesses public concerns for
service costs and effectiveness, respects client involvement in service proce-
dures, and recognizes the needs of the work place, where the fate of innova-
tion will ultimately be decided. This is not likely to occur in this society of
protected interests unless this social and political movement brings the pri-
ority of more humane service provision to the forefront of concern. If such
concern is likely to be evident in the future it will be because place by place
and issue by issue, people-effectively demand respect for themselves and
their proper claims on government, while at the same time they are able to
explore ways to support street-level bureaucracies in their struggle to do a
decent job under adverse circumstances.

211
 CHAPTER 14

On Managing Street-Level
Bureaucracy

Street-level bureaucrats, by definition, have an autonomous core. In a lim-

ited sense, they are the authors of the policies that are finally delivered. But
if the public wants to affect public service policy delivery, it must look not to
the behavior of individual workers but to managers and policy makers.
Throughout the book, particularly in chapter 11, I emphasize the various
ways that street-level bureaucrats are able to undermine or evade efforts to
bring them to account. In this new concluding chapter, written especially for
the thirtieth anniversary of publication, I gather a number of thoughts on
managing street-level workers to bring public services in line with expecta-
tions established, ultimately, through the political process. In doing so, I
comment selectively on developments in the field that inform our under-
standing of the key issues.
I start by treating the origins of the concept of street-level bureaucracy. I
review how the political environment has changed and why it is particularly
timely to learn how teachers, police officers, social workers, and others may
be managed to achieve greater policy responsiveness.

An Evolving Policy Environment for Street-Level Bureaucracy

Street-Level Bureaucracy was written at a time when the United States was
adjusting to the civil rights movement for social and political equality. It was

212
 On Managin g Street-L evel Bureauc racy
written as African Americans and their allies sought to bring into reality the
promises of equal rights. In the early 1970s, when the structure of the book
was being worked out, the ripples of the civil rights movement included con-
flicts over schools, the police, and paternalistic welfare administration. These
institutions were the places people encounter ed or confronted government,
the places unequal treatment would be manifest. In general, they were pop-
ulated by white workers and administrators who became the focus of conten-
tious race relations.
Since the early 1g6os I had been personally and professionally committed
to advancing the objectives of the civil rights movement, as I understood
them. But this did not include sympathy for characterizations of public sec-
tor workers as indifferent to the experiences of the people they dealt with. A
crude racism could not explain on-the-job behavior. Nor did the many inter-
views and observations of street-level bureaucrat s my collaborators and I
conducted support the proposition that personal attitudes or indifference of
street_-level workers could explain the experiences of citizens with these
agencies.
I wanted to understand the performan ce of critical public services that
were either at the heart of community conflict or were central to providing
for citizens in need-the panoply of institutions that focus on citizen wel-
fare: income supports, housing provision, social services, and the justice sys-
tem. I came to believe that the experiences of citizens with public services
were the predictabl e consequen ces of the structure of particular kinds of
publics sector work. This belief was based on the recognition that the work
of police officers and then, I realized, other frontline workers, consisted of
layers of coping mechanisms that guided them through complex interactions
with the people they served, and allowed them to salvage at least a portion of
their work as authentic and worthwhile. My primary concern became recon-
ciling the human dimension of street-level positions with the often problem-
atic outcomes of the systems in which they worked.
Another influence was the emergence in political science of studies that
analyzed what actually happens when public policies are enacted-w ork that
later became known as implement ation studies. It was only in the mid-1g6os
that political scientists began to focus on the now-obvious point that the pas-
sage of laws and the promulgation of regulations did not fully encompass the
policy-making process. On the contrary, as we now take for granted, policy is
only finally made when laws or regulations are fully implement ed, through
subsequen t processes that cascade from the initial declarations.
At least initially, studies of implement ation were part of a discourse that
sought to confront the contradictions between American ideals and Ameri-

213
THE FUTURE

can policy in practice. In 1968, as the issues contested in the streets began to
engage academics, James Davis and Kenneth Dolbeare published a study of
the Selective Service System that asked whether the military draft fairly ap-
portioned risk among potential recruits, as it was supposed to do. 1 They
found that the implementation of the military draft laws led to systematic
overrepresentation of young men from lower socioeconomic backgrounds,
despite elaborate propaganda to the contrary. Studies followed that sought to
understand why substantial plans to enhance economic development came
to grief in Washington, D.C., 2 and Oakland, Califomia. 3 These early studies
were products of a broad skepticism about the capacity of government
shared by many elements in society in the decade deeply influenced by the
"discovery of poverty" amidst plenty, the civil rights movement, ghetto upris-
ings, and the antiwar movement.
Imbedded in the early works on implementation is an implicit critique of
American democracy and an appeal to close the gap between public prom-
ises and performance. Thirty years later, concerns about access to public ser-
vices, and of the relationship between public promises and policy implemen-
tation, continue to be valid. But they are joined by another compelling issue.
We now live in an age in which support for government has been eroded,
and the very purpose of government is deeply contested. The times require
advancing an understanding of the critical roles of government on which
successful modem societies depend: social welfare provision, environmental
and consumer protection, public goods such as schools and highways, and
planning and investing for the future. Appreciating the complexity of public
services through which citizens engage government on a regular basis should
make a significant contribution to this agenda.
The new reality results from several trends that have been well-rehearsed
elsewhere but bear mentioning here. For one thing, today we confront limits
that were hardly on the horizon thirty years ago. In the post-World War II
period, an era of "easy financing" facilitated the expansion of public pro-
grams with only limited concern for whether revenues could be found to
support them. 4 That era is long over. In state governments that by law can-
not run deficits, public budgeting has always been about playing zero-sum
games. Now this is the case at the federal level as well, as federal officials are
under increasing pressure to match any new expenditure with a correspond-
ing budget cut or tax increase.
Another major development has been the sharpening of the tools of gov-
ernment and the invention of new tools. In contrast to earlier times, when
the distinction between governmental and nongovernmental was reasonably
clear, public policies today are coproduced with individuals through tax

214
 On Managi ng Street-L evel Bureau cracy
breaks, credits and vouchers; and with private organizations through tax ex-
penditure s, subsidies, public-private partnersh ips, and contracti ng for ser-
vices.
Separate from these developm ents but related to them are two additional
trends that shape the future of street-lev el bureaucracy. The rise and persis-
tent strength of the conservative perspecti ve on public policy, through aca-
demic and scholarly outlets as well as the pronounc ements of political lead-
ers, has promoted the view that governm ent is intrinsically inefficient and
inherentl y wasteful. According to this perspecti ve, governm ent should be
small and taxes low. In addition, markets and market mechanisms, in con-
trast to governm ent, are said to produce optimal outcomes and lead to inno-
vation and lower costs.
In hindsight I see that in the earlier edition of Street-Level Bureaucracy I
took the existence of governm ent and critical public services for granted, as
most comment ators still do today. I assumed that governm ent services,
though they might change to some degree, were fundamen tally enduring in
their basic structure and support. With the passage of time those assump-
tions seem more tenuous, for several reasons:
1. Contracting for services and other new approaches to service delivery,
which app~ar to introduce market-like mechanisms into public systems,
have been hastened onto the public stage by the conservative challenges to
existing forms of service delivery.
z. At the national and particularly the state and local level, where education,
policing, income supports, and other social policies are administered, gov-
ernments are more or less in permanent fiscal crises as they constantly
struggle to match revenues with the cost of meeting existing obligations.
5

3 Implacable criticism of intrinsic government capacity, when combined

with the public's historic reservations about government, lead to height-
ened attention to government effectiveness.

CONTRA CTING FOR SERVICE S

The widespre ad governm ent use of contracti ng for public services with
nonprofit organizations illustrates some of the challenges in the converge nce
of these trends. In the last several decades, American governm ents have
abandone d direct service provision, such as state homes for delinque nt
youth, and substitut ed instead decentral ized service provision through con-
tracting with nonprofit service agencies. Governm ents subsidize communi ty-
based organizations providing a range of social services, such as shelters for
battered women and supervision of mentally ill homeless people who can

215
THE FUTURE

live independently with assistance. By making funding available, they have

spurred the emergence of flexible and innovative nonprofit organizations by
building into the design of work training and other policies the assumption
that they would be delivered by private contractors. Contracting for services
is now the default approach in many areas, and a new kind of street-level
workforce has emerged to serve citizens in the social services, welfare place-
ments, mental health facilities, and many other service areas.
Although they do not directly work for government, we can expect these
workers to fit the street-level bureaucracy profile, for several reasons. As the
contracting regime has gone from essentially underwriting the policies of
traditional private organizations, such as service agencies, to demanding that
the contracting agencies conform to high standards of accountability, work
patterns on the public side and the private side converge. The controls, per-
formance measures, and agency review procedures imposed on private agen-
cies by public authorities have become increasingly rigorous, tending to
drive out whatever differences in the treatment of clients attributable to
public or private status that might at one time have prevailed. 6
Also, in many street-level bureaucracies, workers' perspectives strongly
reflect professional rather than administrative norms. Thus a social worker in
a contracting agency may process clients very much like a counterpart em-
ployed by a state agency. In some jurisdictions, such as Alameda, California,
indigent defendants in criminal cases may be served by attorneys who work
for local government. In others, such as Brooklyn, New York, they would be
served by attorneys who work for nonprofit agencies. Attorney practices
might vary, but it would be surprising if the variations reflected their status
as employees of government or nonprofit entities. 7
Every public program administered by private organizations under con-
tract is in essence a government program. However, though the contracting
regime has facilitated the growth and greater flexibility of the public sector
by linking it with private community groups, these developments are rarely
treated as public achievements. The transformation of the nonprofit service
sector into what amounts to an arm of government is not credited as an
achievement of government, acknowledged by the sector, or claimed by pub-
lic officials.

FISCAL STRINGENCY
Because state governments cannot run deficits and their tax revenues vary
with employment, wages, and consumer spending, they are subject to budget
fluctuations as revenues rise and fall with the business cycle. Consequently,
they periodically need to cut budgets as revenues decline in economic down-

216
 On Managing Street-Level Bureaucracy
turns. Over the years, as the anti-tax, small-government perspective has taken
hold, state governments have been thrown into more or less permanent bud-
get crises. They have legislated spending limits and passed laws making it dif-
ficult to raise taxes, or have just cut taxes when there have been surpluses,
and cut programs when revenues declined. 8
It has been in the context of continual pressure on limiting the growth of
their budgets in relation to need that the states began to experience the
Great Recession in 2008. One year into the recession, states overall have had
to cut almost one out of every four dollars in their budgets, with more cuts
anticipated for the following year. 9 Such cuts take enormous tolls on the ser-
vices that provide for or protect people in need through street-level bureau-
cracies.
In Ohio, to take one set of examples, a series of tax reductions and a re-
fusal to raise taxes in a recession led in 2009 to the elimination of the state's
early childhood education initiative, and to deep cuts in subsidized child-
care, mental health services, and protective services for children and adults. 10
In October 2oog, the eminent economist Paul Krugman calculated that over
the previous five months the United States had fired more than 143,000
teachers in reaction to the recession, with undoubtedly tens of thousands
more on the block as state revenues continue to be in crisis. 11
No politician wants to reduce the number of teachers, and few want to
reduce the number of other public employees-recreati on workers, nursing
home attendants, and the like. These people work in services of which the
public generally approves. 12 Instead, politicians target the size of govern-
ment overall, making cuts in programs the inevitable consequence of re-
duced revenues. Writing about public benefits at the national level, political
scientist Paul Pierson observes this dynamic: "There is a fundamental asym-
metry between the organized advocates of public spending, who favor par-
ticular governmental initiatives, and their opponents, who (at least rhetori-
cally) criticize government spending in general and on principle." 13

CRITICISM OF GOVERNMENT
Casting a long shadow over the provision of public services is the rise and
influence of the conservative movement, supported by well-funded think
tanks and university centers. Since the 1970s and accelerating in influence
during the Reagan era, conservative theorists and political figures have re-
lentlessly promoted a set of linked ideas to support their version of a better
society: small government, low taxes, reliance on individual initiative (except
for subsidies and tax breaks for favored interests), and reliance on market
mechanisms to achieve efficiency and innovation. Over these years the con-

217
THE FUTURE

servative perspective has contributed to the development of the new tools of

government that include vouchers, contracting for services, and reducing the
range and capacity of public institutions. More important, the conservative
world view has greatly amplified resistance to public institutions as such.
There has always been an antigovernment streak in the American psyche,
and properly so. The country was founded in opposition to an imperial
power, and its constitution was designed in part to limit government's reach.
Moreover, the complex narrative of American history includes government-
sponsored episodes of great cruelty, the tolerance of unjust private behavior,
and the use of public authority to advance the ends of privileged sectors.
However, starting with the New Deal of the 1930s, and despite inconsis-
tencies, government has become the primary instrument for achieving
greater equality and offering shelter from the insecurities generated by peo-
ple's inability to participate fully in the market economy. Government also
establishes the basis for a sound economy by creating the framework under
which businesses operate on a competitive basis without unacceptable costs
to individuals' well-being and the environment. It is this broad understand-
ing of the social purposes government serves today that bear the brunt of the
conservative critique.
In principle, a debate between defenders and critics of an expansive state
can be healthy. That debate is rarely engaged, however. True, particular pol-
icy issues may be roundly debated, but the defense or advocacy of particular
policies is not the same as defending the role of government as such. One
reason government is in disrepute today is that it has many critics but few
active champions or defenders. The critics include not just ideologues and
sincere believers in the view that government is intrinsically flawed or that
taxes are too high, but as well advocates for various interests who agree with
government's objectives but believe it can and should do more. Advocates
with an interest in an expansive government role typically focus on the short-
comings of the sector in which they take an interest, rarely pausing to note
its accomplishments. If they defend anything at all, they defend the pro-
grams they support, but not the set of collective interests of which their fa-
vored programs are a part.
Another source of criticism unredeemed by balancing positive commen-
tary is the tendency of communications media to report on government fail-
ures but ignore government successes. This ensures that when government
comes to public attention it is in a negative light. Moreover, elected officials
rarely praise government, possibly because they fear being regarded as self-
serving. They almost never draw attention to government's essential role.
With conservatives opposing government programs in principle, liberals op-

218
 On Managing Street-Level Bureaucracy
posing them in practice, and news about them focusing only on their failings,
it is hardly a wonder that the reputation of government is low.
Of course, even democratic governments, like all human endeavors, are
sometimes flawed, just as some churches are run by people of questionable
morals, and some corporate leaders bring their companies to bankruptcy or
breach the public trust. But of the major institutions in society, only govern-
ment sustains attacks without effective rebuttal or even measured assess-
ment of the charges.
The good news, I suppose, is that public programs typically enjoy popular
support and are relatively stable over time. New programs are introduced as
the range of problems deemed to require public action expands, and as new
populations are included in the broad consensus of who deserves assistance.
Although recent decades have witnessed vigorous efforts to cut back the
state, the basic composition of the broad social compact remains intact. 14
Unfortunately, although members of the public consistently affirm their sup-
port of the programs government sponsors, their initial reaction to govern-
ment in general indicates broad acceptance of the antigovernment perspec-
tive. A striking illustration of this was revealed in a 1997 Kaiser/Harvard poll.
When asked their opinion about a health plan hotline, 70 percent were
strongly in favor. The number dropped precipitously, to 43 percent, when
the question asked about a government hotline. 15
The shift in the political discourse opens several new paths to what, how,
and whether government will respond to social problems and closes off oth-
ers. The strong antigovernment perspective throws advocates for public ser-
vices constantly on the defensive. Drawing the skirmish line in a public de-
bate at whether something is "governmental" at the least ensures that policy
debates will be ideological (what are the universal principles and values in
terms of which the policy should be assessed?) rather than pragmatic (what
are the best ways to solve this particular problem?).
One reason public managers have been put on the defensive is recent cri-
tiques of government effectiveness that draw on a narrow reading of busi-
ness practices. Demands for accountability often seem to be attacking gov-
ernment programs even as they profess interest in improving them. To
paraphrase Murray Edelman, the insightful guide to politics as symbolic ac-
tion, to insist that one needs greater accountability is also to caution that ac-
countability is lacking. 16
A case in point is the movement to demand that social programs be sub-
ject to the highest evaluation standards, of which controlled experiments
with randomized assignment of subjects is the gold standard. Every policy
analyst recognizes that such research, when appropriate and properly exe-

219
THE FUTURE

cuted, yields unimpeachable results. Within a certain range, the approach

has unquestionable valueP Such research is often not appropriate, however.
Experimental design with random assignment cannot carry the weight of as-
sessing public programs across the board.
Applied as adherents appear to advocate, one result would be to recom-
mend or insist on random assignment experiments even when they are inap-
propriate. For example, if the experiment involves mediation of street-level
bureaucrats in the provision of a benefit or service, the results would likely
be affected in unknown ways by variable implementation of the program. In
such an example, a modest finding might reflect that the intervention was
not particularly useful, or that a useful intervention was erratically imple-
mented.
A second result would be to discount as unproven promising administra-
tive reforms that cannot meet the test of random assignment experimenta-
tion. A third result would be to discount other approaches to sound policy
making that are based on the best available evidence, but also on judgment,
experience, and expert insight. At the symbolic level, insistence on policies
that can meet the test of random assignment radiates the message that other
ways of knowing and deciding are weak and illegitimate.
Governments and foundations, seeking to emulate what they believe are
effective business practices, increasingly insist on evidence standards that
are often inconsistent with needs of public and nonprofit agencies that help
individuals and work with communities. At a certain level, these practices
make eminent sense. At another, they seem intent on creating an evaluation
environment that will markedly limit reasoning about public policy. 18
It is extremely difficult to dislodge the antigovernment perspective. Nei-
ther the debacle resulting from the government's slow and uncoordinated
response to Hurricane Katrina, nor the recent turmoil resulting from the
extraordinary failures of financial regulation, resulted in more than a short-
term boost in public appreciation of the importance of investing in public
institutions. To be sure, President Obama has regularly reminded the public
of the importance of government, providing welcome relief from the nega-
tive rhetoric of the past. 19 But to judge from continued opposition to govern-
ment intervention in the health insurance industry and financial institu-
tions-the pivotal policy issues in the United States in zoog-it would seem
safe to conclude that antigovernment perspectives are an enduring feature
of the landscape.
Recognizing that the early twenty-first century is characterized by deep
skepticism about government, efforts to improve government performance
take on new meaning. Improving schools or the welfare system or policing

220
 On Managing Street-Level Bureaucracy
are not just narrow matters of achieving more effective public services at the
appropriate cost. They may also be understood as contributing to a more
substantial agenda in which government, by improving its public services
across all the divides of race, ethnicity, and class, is perceived as fair and
trustworthy.
In the next section I treat the central question of whether and how street-
level bureaucrats can be managed when their work is defined as exercising
discretion. I then take up the challenge of improving the capacity of street-
level bureaucrats to make sound judgments when relatively standard ap-
proaches to management are insufficient in achieving public purposes.

Shaping Street-Level Bureaucrats' Performance

DO STREET-LEVEL BUREAUCRATS "MAKE" POLICY?

The logic of democracy requires that citizens and lawmakers alike have a
high degree of confidence that enactment of a law will be followed by reli-
able implementation. As we know, however, accountability is complicated in
street-level bureaucracies because workers in these organizations by defini-
tion cannot be fully controlled. Agency policies are not entirely compre-
hended by their authorizing statutes, regulations, budget allocations, and of-
ficial procedures. Nor can discretion in these jobs easily be eliminated. From
disability assessments of people with lower back pain and other ambiguous
conditions, to police encounters in which the right thing to do must be con-
structed on the spot, discretion is required of those who deliver policy. 20
Nonetheless, although street-level bureaucrats are widely expected to ex-
ercise discretion in the course of their work, managers must design ways to
insure accountability at the street level. Earlier in this volume I observed
that the actions of teachers, police officers, or welfare workers "become, or
add up to, agency policy," and that their actions effectively "become" the
public policies they carry out."21 For a great many of the readers of the origi-
nal edition, these conclusions were the primary and sometimes the only les-
son of Street-Level Bureaucracy.
This interpretation is clearly too limited. Street-level bureaucrats may in-
deed "make" policy in the sense that their separate discretionary and un-
sanctioned behaviors add up to patterned agency behavior overall. But they
do so only in the context of broad policy structures of which their decisions
are a part. Street-level bureaucrats do not articulate core objectives or them-
selves develop mechanisms to achieve them. For any given public agency or

221
THE FUTURE

any policy reform, we need to look into the entire policy environment in
which street-level bureaucrats function.
The transformation in 1996 of the basic American welfare program was
obviously a critical policy decision that took place far beyond the interview
rooms of local welfare offices. The new program ended welfare as an entitle-
ment and made eligibility depend on work effort and job-seeking. Street-
level bureaucrats, who only a few weeks earlier were focusing on monitoring
the accuracy of applications and applying eligibility rules, were now required
to learn the vocabulary and ways of a system that expected clients to look for
work and find it. And, ideally, they would have to learn how to apply that
vocabulary to welfare applicants as they presented themselves, and strategize
about the most appropriate path for clients to enter the workforce. These
changes reflected policy making at higher levels. In addition, options avail-
able to workers in processing their caseloads were heavily influenced by sev-
eral factors.

1.Work opportunities in the welfare agency's catchment area, including the

kinds of jobs available, entry level wages, and clients' access to public
transportation.
2. The availability of child-care services and health care (particularly for sin-
gle parents whose reliability at work depends on the health of their chil-
dren).
3 Client training and placement opportunities.

These (and other) factors shape the decision making of today's welfare
workers. As Evelyn Brodkin observes, "the quality of choices about help
[available to workers] depend[s] in part upon the 'helping resources' to
which street-level bureaucrats have access." 22 As always, they depend on
caseloads and other work circumstances that impinge on conditions of work
in the welfare system. But since welfare reform they also depend on some
new factors-especially the local employment situation-that are beyond
their control.
Consider the New York Times report in 2009 that "more people were be-
ing turned down for welfare benefits because of overly stringent or confus-
ing requirements."23 It may be that welfare workers were exercising exces-
sive discretion in discharging their responsibilities, but it also could be that
the rules in New York really were overly stringent or confusing.
In short, in the new work and welfare systems, workers categorize appli-
cants and deploy available resources according to routines and simplifica-
tions they develop, even as they exercise judgment and make unsanctioned

222
 On Managing Street-Level Bureaucracy
responses. The content of these discretionary acts will not be random, and
they will not only be related to the coping needs of the workers. On the con-
trary, they will be structured by the choices available to the workers as op-
tions (such as local funding for daycare) provided by the policies.
To understand worker behavior we need to pay much more attention to
the overall structure of the policies. A study of nurses' preferences for public
versus private hospitals in Australia concluded that the shortage of nurses in
public hospitals might be attributed to their having less autonomy than their
counterparts in private hospitals. The hypothesis is plausible, but other im-
portant factors, which the study did not investigate, might have accounted
for the nurses' job preferences, including whether the private hospital nurses
were paid higher wages, worked better or fewer hours, served a more conge-
nial clientele, or had lower caseloads.

CAN STREET-LEVEL BUREAUCRATS BE MANAGED?

Over the years, some readers of Street-Level Bureaucracy have also con-
cluded that street-level bureaucrats cannot be managed. It is true that ear-
lier in the book I elaborate on the resources lower-level workers have that
allow them to resist directives from supervisors. Managers of street-level bu-
reaucrats are limited in the ways they can exercise authority, but thirty years
of experience have shown that managers can surely narrow the gap between
the performance of street-level bureaucrats and the desired policy results.
One approach is to minimize the mediating role of street-level workers in
citizens' accessing services. 24 Administrators can radically reduce discretion
by narrowing the job, as in the case of welfare departments converting the
job of social worker from counselor to greater emphasis on the clerical. 25 In
Britain in the 1ggos, introduction of principles of the "new public manage-
ment" resulted in administrative reforms that narrowed workers' discretion
but also focused the objectives of the policy more sharply. In the case of
teachers, school managers gained greater control by emphasizing clearer
standards. In the case of geriatric social workers, policy makers placed
greater emphasis on social workers' roles in structuring services for the el-
derly as they reduced their focus on counseling. Predictably, the reforms
shifted and constrained discretion, but did not eliminate it. Workers often
reported positive results because the requirements of the jobs were clearer,
although many were dismayed by what they considered a loss of status and
authority. 26
Partly to reduce workers' discretion when unnecessary for effective agency
functioning, but mostly to increase productivity and reduce costs, people-

223
THE FUTURE

processing agencies have more recently devoted enormous resources to au-

tomated information gathering and intake functions, a trend accelerated by
the Internet and the widespread availability of computers. The application of
information technology can lead to extraordinary productivity gains. To take
one example, the much maligned departments of motor vehicles, the agen-
cies that have been poster children for government inefficiency, have ex-
panded services, reduced waiting times remarkably, and kept costs down
through data processing innovations and the use of the Internet. In Virginia
in zooS, for example, the DMV reduced the average waiting time in cus-
tomer service centers over the year from a tolerable thirteen minutes to
eight. Although the number of drivers, vehicles, and agency responsibilities
have increased steadily since 2002, the D MV workforce in Virginia declined
by 3 percent. 27
Like any tool, extensive use of information technology can be well-used or
used irresponsibly. When the state of Texas attempted to reduce the number
of welfare workers by contracting with a private provider to receive applica-
tions online, the misguided innovation deprived people of coverage. Enroll-
ment in the Children's Health Insurance Program, for example, declined by
127,000 children in a five-month period. 28 This is a dramatic case of a gen-
eral issue with streamlining intake through self-service: the costs for citizens
and potential service claimants of being left on their own go unrecorded,
and thus do not register as agency failures. 29
Automated information systems do not eliminate concerns over respon-
siveness-treati ng people as individuals who may present unique circum-
stances-even in innovations such as providing or seeking fairly straightfor-
ward information from citizens. In 1997 the Australian government instituted
a computerized system to provide assessment and referral functions to the
unemployed. The scheme was touted as offering more finely tuned assess-
ments about individual capacity and potential, but was also intended to en-
force harsh financial sanctions that frontline workers could not be trusted to
levy. According to researcher Greg Marston, the system failed in helping
job-seekers with fine-grained decisions, and the task fell to the cadre of
counseling and training organizations charged with actually helping people
find work. 30
Other concerns with automated service systems include ensuring that da-
tabases are kept up despite pressures to reduce costs, and maintaining sys-
tems even when increased access to an agency may result in higher demand
to which it would be costly to respond. 31 Information systems must make
sure that assistance is still available to those who do not have computers or

224
 On Managing Street-Level Bureaucracy
cannot use them, and that language barriers and other issues of citizen ac-
cess are addressed.

RESTRUCTURING INCENTIVES AND SANCTIONS

Aside from simply reorganizing the problem of street-level discretion by
reducing it, the more significant way managers influence street-level bureau-
crat discretion is by restructuring worker incentives and sanctions. In such
restructuring, street-level bureaucrats' actions may still "add up" to the pol-
icy the organization produces, but the policy in fact conforms better to man-
agers' expectations. Essentially, managers increase the probability that out-
comes will shift in the preferred direction. 32
Five examples illustrate this approach. Of these, the first four recount ef-
forts to alter the behavior of street-level bureaucrats toward clients, in the
first three reducing workers' generosity, but in the fourth expanding their
helpfulness.
Diagnostic related groups. In an attempt to control medical costs, since
1983 the Medicare program has required hospitals to assign most new pa-
tients to diagnostic related groups (DRGs). The DRG system provides a ba-
sis for standardizing treatment and controlling cost by assigning a previously
calculated payment level for the treatment of the diagnosed condition rather
than paying for every procedure. Hospitals have a strong incentive to elimi-
nate procedures that might result in costs that exceed the reimbursement
rate. Although DRGs can be and are gamed (by assigning ancillary diagnoses
to cover unexpected costs, for example), this and similar approaches have
gained wide approval in the health care field as a way to rein in expensive
outliers without intruding on doctors' medical decision making. In essence,
the use of DRGs leads to convergence around a norm of treatment and pay-
ment, but without appearing to tell doctors what to do or threatening the
sacred doctor-patient relationship.
Quality control in welfare administration. Every substantial enterprise,
public or private, is concerned with maintaining standards and insuring
against the degradation of products or processes. In welfare administration,
quality control essentially means that welfare agencies, down to the office
level, collect information to ensure that they correctly administer state and
federal policies. In the 1970s, but particularly in the early 1g8os, federal wel-
fare officials sought to use the quality control system not just to ensure con-
formity with the law, but also to curb what they regarded as excessive wel-
fare expenditures. The mechanism for reducing expenditures through quality
control was engagingly simple. The federal government would penalize

225
THE FUTURE

states for excessive errors in wrongful spending-adm itting ineligible people

to the rolls, or authorizing excessive payments. But in determining sanctions
they would not count errors of stringency-tha t is, errors in which workers
denied eligible applicants, or authorized payments that were too low.
As a result, state campaigns against error focused exclusively on inducing
workers to be tougher, and to lean in the direction of refusing claims if there
were any uncertainty. Ambiguities in the law were resolved in ways unfavor-
able to clients. Recipients already on the rolls were burdened with additional
behavioral costs as demonstrating compliance with eligibility requirements
became more onerous. Quality control appears to be an unobjectionable
neutral administrative mechanism. As designed in this case, however, it had
distinct distributional consequences. 33
Reining in disability insurance determinations. Excessively liberal re-
sults in the Social Security Disability Insurance (SSDI) program were the
subject of similar concerns in the same period. Federal administrators be-
lieved that many of the independent administrative law judges (ALJs) were
not strictly applying the SSDI rules. In response they mounted a campaign
that focused on the rulings of those ALJs whose records suggested that they
tended to favor claimants.
Among the tactics to redirect the tendencies of judges deemed to rule ex-
cessively in favor of claimants were developing a protocol that tried to rou-
tinize the decision process; tracking and publicizing individual judges' rever-
sal rates to draw attention to those who were deemed to be excessively
generous, and then 'counseling' those judges who deviated most from the
norm; subjecting judges' decisions to quality control protocols; and subject-
ing judges' decisions to a time limit, which constrained judges from gather-
ing additional evidence that might be favorable to a claimant's case. 34 Again,
these administrative changes focused on changing the choices ALJs made
without challenging their right to make judgments on any particular present-
ing case.
Transforming welfare worker orientations. The three examples given il-
lustrate administrators' efforts to increase the probability that discretionary
judgments would be made to withhold benefits to potential recipients.
Something of a contrary example is provided by the transformation of the
Massachusetts Department of Welfare in the 1g8os from an agency primarily
focused on income support activities and monitoring the eligibility of recipi-
ents for cash payments, to one devoted to placing welfare recipients in jobs.
In 1983, the new welfare commissioner, Charles Atkins, adopted the then-
unprecedented objective of placing recipients in jobs or preparing them for
work with the goal of helping them get out of poverty. Employment and
 On Managing Street-Level Bureaucracy
Training Choices (ET) was one of the first programs in the nation to take
seriously the importance of placing welfare recipients in jobs-a prescription
that would dominate discussions of public welfare over the next decades.
The ET program would offer welfare recipients three main services: assess-
ment and career planning, training and placement, and further education in
support of work preparation. Because the program would be voluntary, case
workers would have to persuade recipients to enroll, counsel them as to the
best options, and encourage them to persist if the experience was not satis-
factory at first. These were the same case workers who just months before
were focused exclusively on assessing eligibility and determining welfare
award levels.
It was no easy matter to induce welfare recipients with low self-esteem,
little work experience, and few employment-related skills to believe that the
new program could help them succeed and was worth the personal and fi-
nancial risks. Nor was it an easy matter to persuade DPW workers to believe
in and adopt the new objectives. Workers would have to establish rapport
with recipients, engage them in discussion of matters not directly related to
the recipients' primary concerns of getting on welfare, and persuade them to
investigate the program's options. This persuasion would have to be done by
social workers who were not trained for such work and who, a short while
ago, had been focused on an entirely different set of priorities.
In 1987, both the department and the independent Massachusetts Tax-
payers' Foundation agreed-though they differed in the details-that ET
had placed tens of thousands of people in jobs, reduced welfare rolls, and
saved taxpayer money. Although some critics regarded the ET claims as un-
proven because the state resisted, on sound grounds, subjecting welfare re-
cipients to an evaluation with random assignment, there should be little
question about the transformation of the state's welfare staff. 35
In his study of ET, Robert Behn summarizes how managers radically re-
shaped the behavior, motivation, and production of these street-level workers. 36
1. They established a clear mission for the agency, and developed a wide
range of vehicles and events to communicate it throughout the ranks.
2. They set specific goals for individuals and for each work unit leaving no
doubt about the agency's highest priorities.
3 They monitored performance and rewarded workers' success at the indi-
vidual and unit level.
4 They were personally involved in monitoring agency performance.

ET established a new mission for the welfare department and recruited its
workers to support the mission by relentlessly promoting indicators of sue-

227
THE FUTUR E

cess, keeping track of performance , and conveying to workers that they were
critical to the new mission of the agency. 37
CompSTAT. A successful and widely imitated initiative that appears to
embody these principles is CompSTAT, the police administrative reform be-
gun during the administration of Commission er William Bratton in New
York City in 1994. 38 CompSTAT takes advantage of high quality data collec-
tion and reporting of crime statistics at the precinct level to hold precinct
commanders accountable. Precinct commanders , like the welfare adminis-
trators subject to quality control initiatives mentioned earlier, in tum engage
police officers to achieve the desired results.
In monthly meetings with the Police Department' s high command, ad-
ministrators require all precinct commanders to be ready to defend their
performance and explain trends in their districts that require attention. If
there has been a spike in reported burglaries, for example, the commander
can expect to be interrogated on why this trend is occurring, and what he or
she intends to do about it. The system works when precinct commanders ,
with clarity about their priorities based on the precinct data, and urgency
deriving from the monthly meetings, design strategies to improve the perfor-
mance of their units. At a certain level the critical elements present in the
ET case are present here as well: clarity of mission, close tracking of perfor-
mance, and keen involvement of the organization's leadership.
In evaluating these reforms a key question is whether focusing on a spe-
cific aspect of the mission pulls street-level bureaucrats away from the full
range of their responsibilities to a detrimental degree. Does monitoring and
evaluating performance on a particular measure improve job performance
overall? Or does focusing on a single dimension of the job subtly change the
job itself in negative ways?
In CompSTAT, police are not relieved of responsibility for law enforce-
ment outside of departmenta l priorities, and in any case, if new crime trends
emerge, they also can be targeted. In the ET experience, job placement was
only one of many departmenta l priorities that received attention through
monitoring and feedback from the high command. For example, in ET
maintaining a good record on quality control remained a specified agency
priority even though the first priority was recipient employment . CompSTAT
has been widely emulated in other police departments , and its principles
have been widely adopted in efforts to improve accountability in other agen-
cies and-in the cases of Baltimore and Washington State-gover nment-
wide initiatives. As with any replications, success in these efforts depends on
how faithfully the model is actually reproduced. 39
These reforms of street-level bureaucracy seek to harness measuremen t
 On Managing Street-Level Bureaucracy
of performance with appropriate sanctions and rewards. Their success ap-
pears to depend upon the alignment of the new approach with the core val-
ues of the agency, the clarity of the mission, and the extent to which street-
level bureaucrats can in fact improve their performance in pursuit of the
mission. 40 These examples demonstrate that managers can influence the be-
havior of street-level bureaucrats by strategically choosing the data they col-
lect about workers' performance, and the incentives with which they respond
to the data.

Investing in Street-Level Bureaucrats

A paradox of public service provision in democratic societies is that policies

must be administered fairly; similarly situated people must be treated alike.
And yet, as James Q. Wilson emphasized many years ago, we also want our
public services to be responsive to the presenting case. 41 Although respon-
siveness to each case in practice would be a nightmare, at least in principle
we want public services to recognize important and relevant differences. In
pursuing responsiveness, managing street-level bureaucracy by manipulating
measures, incentives, and sanctions is of relatively little use.
Public service bureaucracies must emphasize responsiveness in two cir-
cumstances. First, studies of street-level bureaucracies and everyday experi-
ence remind us that we want and expect public officials to be flexible and
helpful. Consider the following examples.
1. The social worker, subverting strict rules, retains a welfare recipient on the
rolls while she acquires necessary documentation.
z. An intake worker coaches the job training applicant on how to present his
case in order to qualify for a benefit.
3 The hospital team tags a patient with a reported medical condition that al-
lows her to stay in the hospital a day or two longer while team members
shape a better discharge plan.
4 The teacher spends extra time with a new student.
5 The police officer accompanies the speeding motorist to the hospital as the
motorist's wife is in the throes of childbirth.
In all these cases, the caring displayed by the public employee captures
our imagination and confirms our belief"that public structures are capable of
treating people as individuals and not just as units to be processed.
When those with authority go out of their way to respond generously to

229
THE FUTURE

people in need despite rules to the contrary, they join what sociologist Lisa
Dodson smartly calls "the moral underground."42 Members of the moral un-
derground do not blow up rail lines. What they do is ignore rules that they
are supposed to enforce but seem excessively rigid, in order to help people
in need, at little overall cost to society.
Acts of flexibility not only benefit the recipient of compassion but also
have larger implications for society. After presenting a case in which a voca-
tional rehabilitation counselor cut through red tape to obtain a computer for
a client, Steven Maynard-Moody and Michael Musheno describe the coun-
selor as engaging "in an act that redeems the state by breaking through the
bureaucratic labyrinth." 43 In addition, they suggest, such an encounter pro-
vides a context in which community values that coexist and sometimes con-
flict with bureaucratic rules find expression. Many administrative systems al-
low for such flexibility under certain circumstances, though flexibility can be
misused and can result in wholesale undercutting of intended policy. Like so
many matters in policy analysis, the proper degree of flexibility is a matter of
balance.
Treating everyone alike is a requisite of building popular trust. But good
rules allow for appropriate exceptions. How, then, can a society tolerate be-
nign exceptions when universalism is so important? Exceptions that favor
people because of their social ties or ethnic backgrounds undermine popular
trust in government's fairness. We cannot specify where the lines should be
drawn, but we can acknowledge that responsiveness within a norm of univer-
salistic rule application can be consistent with fair and effective public ser-
vices, even if it requires departure from strict mechanical accountability. The
exceptions that mark the superior public agency must be those that would
pass a test of deserving popular approval on the basis of the presenting facts
and local values, and are not tainted by implications of favoritism based on
the identities of those favored.
A second circumstance in which public service bureaucracies must em-
phasize responsiveness is when public services requires individual initiative,
the cultivation of experience, and a degree of empathy that cannot be re-
duced to administrative guidelines. Thus there is another '"problem" of
street-level bureaucracy that is entirely different than the one we have con-
sidered up to now: how to insure that people employed by the state to teach,
judge, evaluate, and counsel have the necessary skills, experience, and train-
ing to exercise discretion properly and most effectively. For some street-
level occupations more than others, it is essential that decisions are made
fully in response to the individual case or situation, deploying deep knowl-
edge of the field with commitment and a common sense that cannot be codi-

230
 On Managing Street-Level Bureaucracy
fied. As Michael Hill and Peter Hupe put it, "[e]nhancing street-level discre-
tion may, under certain conditions, be more functional for the implementation
of those policies than curbing it." 44

NO CHILD LEFT BEHIND

These issues are reflected in the drama that has been playing out in the
United States for most of the current decade in educational reform. The
George W. Bush administration's legacy legislation, The No Child Left Be-
hind (NCLB) Act of 2001, aspires to improve educational outcomes for all
children, and particularly to close the achievement gap among young people
of different socioeconomic groups. To accomplish its goals, NCLB requires
all states to test all children in certain grades in math and English and to
sanction schools that fail to make adequate progress. In essence, it is an
enormous exercise in performance measurement, with individual schools as
the units of analysis.
The theory of NCLB is that the reality or the threat of poor performance
ratings and the prospect of severe sanctions for poor school performance will
mobilize school administrators and their communities to improve the
schools. Whether the act has produced measurable progress is disputed, but
there are some indications of improvements in school performance, particu-
larly in schools with high proportions of students from disadvantaged racial
or ethnic backgrounds. Educational reformers now rally behind a substantial
reform scenario rather than reject the new approach outright. 45
One obvious problem with the law, identified by many observers, is that
the emphasis on testing skews the curriculum in the direction of test achieve-
ment to the neglect of other concerns that do not directly contribute to
higher test results. The task of schools has never been exclusively to convey
information. Mass testing does not assess the ability of students to manipu-
late knowledge in ways they will need to in later life, to say nothing of stimu-
lating creativity, coaching young people in how to get along with others, and
cultivating civic virtues, to name only a few of the qualities of mind reflective
of a good education.
There is wide understanding in the research community that "teaching to
the test" has several drawbacks. It drives out of the curriculum subjects on
which students are not being tested, such as art, physical education, and cul-
tural studies. 46 It also distorts and excessively simplifies the subjects that are
tested. More broadly, the tests ignore the demand from all segments of soci-
ety that children "become competent problem-solvers and critical think-
ers."47 In principle, states could improve their tests, but such efforts are
costly, and in any case states have a perverse interest in reducing the rigor of

231
THE FUTURE

the tests in order to ensure that more schools meet the minimum stan-
dards.48
A deeper problem with NCLB, and one that generally confronts perfor-
mance measurement reforms, is the failure to address the strategic question
of how to improve achievement. In principle, performance-based account-
ability leaves open the question of how results are achieved; encouraging
flexibility in achieving goals is more or less the point of these approaches.
However, if the construction of a performance-based regime absorbs exces-
sive resources and the attention of administrators, creating ways to achieve
better results may suffer. This point is particularly relevant to NCLB, given
the substantial evidence that improving the capacity and training of teachers
is one of the most powerful paths to improved educational outcomes. 49
NCLB requires the states to employ "highly qualified teachers," but to do
so schools need only hire educators in core academic areas who are licensed,
hold a bachelor's degree, and have taken courses in their subject area (as evi-
denced, say, by their undergraduate major). This is a decidedly shallow stan-
dard of what constitutes "high qualification" of a good educator. Higher ex-
pectations of teacher training and preparation would yield much better
results. Recommending the approach of the National Board for Professional
Teaching Standards (NBPTS), Mary E. Dilworth and Joseph A. Aguerrebere
suggest that highly qualified teachers must meet the following criteria.
1. They are trained in their subject areas and broadly grounded in the liberal
arts so as to be able to help students make connections across issue areas.
2. They command a variety of approaches to learning and are able to adapt
their approaches to students with different learning needs and styles.
3 They are committed to teaching and the belief that all students can leam. 50

The last is not a small matter and is separable from subject area knowl-
edge and teaching methods. Good teaching depends on dedication and on
the teacher's conviction that students will be able to succeed. Students who
are taught by teachers certified by NBTPS, in 2009 numbering almost
83,000, achieve at a higher rate than those taught by uncertified teachers, as
measured by evaluations in North Carolina, Arizona, and Florida. 51
In principle, all street-level bureaucrats, from health-care workers to po-
lice officers, must be able to assess the presenting situation or client freshly.
Equal treatment may require treating people differently to achieve equal re-
sults, particularly when responding to diverse populations. In the case of
teachers, according to Dilworth and Aguerrebere, this means that teachers
should know their students well, both as individuals and as children from
shared but varying backgrounds. Teachers who apply "culturally responsive

232
 On Managing Street-Level Bureaucracy
pedagogy" will understand, for example, "what types of errors a student who
is learning English is likely to make based on patterns that exist in their na-
tive language."52
Accountability systems with single indicators are magnets for criticism be-
cause public sector services rarely have unitary goals. 53 The challenge for
critics and supporters of No Child Left Behind is whether reducing the mis-
sion of the schools to achievement on the current tests is an acceptable re-
conceptualization of our expectations for schools. A generous assessment of
NCLB is that it has placed achievement of every student at the center of a
national debate, and located the problem at the level of the individual school.
As mentioned, the law is distressingly simplistic on what constitutes a good
education, and it is also more or less silent on how to achieve better results.
Despite substantial criticisms, the law will have proved valuable if future it-
erations lead to support for substantial investments in teacher salaries, work-
ing conditions, standards, and training.

THE CASE OF CHILD PROTECTIVE SERVICES

Another occupation that cries out for greater investment in training and
preparation for encounters with citizens is child protective services. Child
protection workers are in some ways the ultimate street-level bureaucrats.
They exercise police powers in their mandate to remove endangered chil-
dren from their homes, but, in the name of supporting families, are expected
to exercise this authority as infrequently as possible. Although their presence
in a family situation always suggests that they may decide to separate chil-
dren from their families, CPS workers are nonetheless expected to use their
interpersonal skills and agency resources to help families cope with crises.
Like police officers who live with the knowledge that a wrongful shooting
may result in the loss of innocent life, and on a personal level may lead to
severe discipline and public notoriety, CPS workers are vulnerable for the
fateful judgments they must render.
There is no decision more fraught with potential for serious errors of
omission and commission than those concerning whether children should be
removed from their families. Removing a child from a home can be heart-
breaking and often leads to results from which it is hard to recover, even if
the child is returned home after a period of separation. Yet leaving a child in
a home where violence or neglect may occur can be even more dangerous
for the child, and in a different way, for the worker. Rarely are the choices
clear cut.
Yet who is typically called upon to render judgments of such conse-
quence? Here is how Harry Spence, commissioner of the Department of So-

233
THE FUTURE

cial Services in Massachusetts in the early zooos, described the workforce he

oversaw:
We entrust the basic work of ensuring the safety of children to individual social work-
ers, generally between the ages of 23 and 30, with an average tenure in the Depart-
ment of 1 year and 11 months, with one month of training. We ask them, with only
one set of eyes and ears, to observe each of at least 18 families for a few hours each
month, often in circumstances of considerable tension and even danger. We require
that they then predict the future behavior of that family, and make gut-wrenching,
life-shaping decisions on the basis of those scant observations. They live with the
constant knowledge that if they are wrong, a child may die, and they will be fired and
publicly excoriated. 54

The challenge for the Massachusetts Department of Social Services (DSS)

was not to develop accountability mechanisms to secure compliance with
agency objectives. Those already existed in such indicators as the number of
home visits and supplemented with extensive case reporting requirements.
In any event, the operative or tacit performance measure was whether a
worker's cases stayed out of the newspaper.
In considering whether he should take the job, Spence saw in the Massa-
chusetts DSS an ultimate public administration challenge: inexperienced
and poorly trained workers practiced defensive social work and had little
support from management if they erred. To make sound decisions, the work
of DSS had to become better work.
With respect to workers' preparation, Spence invested in worker training,
establishing an institute for protective services workers at a state college
near Boston. Like the approach to teacher training mentioned earlier, he
also emphasized the empathic requirements of the work. Social work was as
much an "art" as a skill set, he wrote, consisting of "an open and sensitive
emotional connection to the family and child," and a "habit of self-reflection"
that were supported by a "community of self-reflective practitioners."55
These observations are notable because in the public mind child protection
workers are not seen as empathetic professionals but come to attention only
when a decision or a nondecision has gone fatally wrong.
Reflecting an understanding that protective service work required re-
thinking the structure of decision making in the field, Spence introduced
two unusual reforms designed to overcome the paralyzing, fear-driven model
of child protection practices.
One innovation was to introduce into the department's routines the medi-
cal model of error prevention. In the past, workers who experienced even
near-disastrous outcomes were afraid to talk about them or seek help for fear

234
 On Managing Street-Level Bureaucracy
of being reprimanded. The consequences of making a mistake were so great
that workers who had committed "near-errors" never discussed them openly
or sought advice about how they might have handled situations differently.
DSS lacked the cultural norm of being able to ask for help or advice.
In response, Spence introduced a version of no-fault case review widely
adopted in U.S. hospitals after being pioneered in the Veteran's Administra-
tion.56 Under this model workers are encouraged to bring up troubling cases
for review. The model rejects the "error and blame" approach to account-
ability in favor of a culture in which it is safe to discuss and acknowledge er-
ror the better to encourage organizational learning. The no-fault model
works by making it safe to bring up problematic situations on the assump-
tions that no sanction will result from bringing up problems in the field, and
that such discussions are the basis for workers' and the organization's con-
tinuous improvement.
The second innovation was to introduce the team approach to case work.
Two workers rather than one would be responsible for each case. This ap-
proach would directly confront a critical imperfection in the existing system,
in which a single worker, with little experience and few resources to cope
with the tensions of the visit, would call on a distressed family. Family mem-
bers were often hostile or unresponsive in these settings, and often withheld
critical information about the functioning of the family.
A team approach might be better. No longer would an individual worker
be the solitary judge making a momentous decision alone. Workers could
share notes and consult with one another on a course of action. Families
were more forthcoming, perhaps because they believed they could appeal to
another worker if one did not seem sympathetic, or because the team ap-
proach conveyed to clients that the agency took their case more seriously. In
any event, the team approach found favor with workers and clients alike.
No outcomes in protective service interventions are certain, particularly
when the philosophy of child protection toggles between removing children
from the home and maintaining family integrity. Children die or are hurt
while families are under protective custody orders, just as they die or are
hurt in families not designated as requiring state intervention. In the new
approach to accountability, Spence promised never to scapegoat workers
when a child died in the care of the department-if the worker had made a
good-faith effort to execute agency policy. He acknowledged that the work
was fraught with danger and uncertainty, and said he would protect workers
when misfortune occurred, as it inevitably would.
This approach to reforming protective services recognizes that human
judgment is essential for effective public policy, and that the central chal-

235
THE FUTURE

lenge for management is to improve workers' capacity to render that judg-

ment dispassionately, as much as possible based on the presenting case, and
as little as possible on extraneous fears or the instinct to take the easy way
out.
The case of protective services is a dramatic one, but the principles are
not unique to this policy arena. They apply, for example, to the case of chil-
dren's mental health services, in which professionals must recommend a
drug regimen and referral to supportive programs with relatively little cer-
tainty about whether the client will respond positively, but with the knowl-
edge that the referral will be consequential for the client and the family or
the placement environment.
Street-level bureaucracies such as child protective services, mental health
assessments, and the police are actors in arenas in which issues of public
trust are particularly mediated. We have some reason to believe that citizen
trust in society is related to the effectiveness of street-level bureaucracy,57
just as we have confidence that one of the components of public trust is fair-
ness in the application of the law and the administration of public policy. 58 In
the case of street-level bureaucracies with episodically high, negative pro-
files, much more attention should be paid to helping citizens appreciate the
complexity of efforts to achieve public purposes and the futility of banishing
all mistakes. As Murray Edelman observed, it is "the routine functioning of
... street-level bureaucracies without too much controversy [that] confers a
reputation for leadership upon their immediate heads and also upon the
highest officials of the regime."59 Our knowledge of the linkages between the
actual effectiveness of public services and support for an expansive public
sector includes considerable speculation, but it is a subject worthy of greater
attention.

Conclusion

In the 1970s, as I noted in chapter 1, some 3 7 million people were em-

ployed in American public schools, most of them teachers. Now that popula-
tion includes more than 53 million teachers. Then, more than half a million
people were employed in police work; now close to 1 million are. Then,
300,000 worked in public welfare; now more than 540,000 do, with many
more working in nonprofit agencies performing welfare functions. 60 These
public sector jobs increase with the population, are relatively impervious to
 On Managing Street-Lev el Bureaucra cy
automation, and will not be relocated overseas. By reasonable societal stan-
dards, they pay well.
These and other public services employing street-level bureaucrats have
weathered the storms associated with racial integration of their members,
and of the communities they serve. Immigration has introduced new cul-
tures, styles, and languages to the streets, classrooms, and waiting areas
where people still receive what the state has to offer or imposes on them.
They have been transformed by new ways of communicat ing and record
keeping. Many have been challenged by new ways of organizing services, in-
cluding vouchers, charters, and contracting with nonprofit agencies.
Ultimately, the work of street-level bureaucracie s comes down to relation-
ships established in single encounters, or over time, between citizens and
the people whose job it is to render services, provide support, or make judg-
ments about how citizens fit the laws and practices of public agencies. Shap-
ing those encounters are the supervisors, managers, and policy makers who
establish the breadth, contours, and particular character of the exchanges.
Behind these scenes are virtual industries of people who share an interest
in improving the effectiveness of particular public services. Each street-level
bureaucracy has its own corps of university faculty, researchers, federal fund-
ing officials, consultants, and interested foundation officers, as well as unions
and professional associations at local, state, and national levels that produce
research and reform agenda, lobby for favorable outcomes, and keep mem-
bers informed. These diverse individuals believe in improving the circum-
stances and performance of street-level occupations.
They also share a paradoxical political environment. The work they per-
form or support is widely accepted and often highly approved, but the public
sector of which they are a part is held in low regard. They experience day-to-
day operations as understaffed, and adequate resources seem perpetually re-
mote. This is the current challenge for advocates of these public services:
how to balance respect for the individual encounter that is at the heart of
street-level service provision, and at the same time negotiate the larger ques-
tions of achieving efficiencies, sustaining adequate revenues, and serving as
witness for the critical role of these public systems in civic life.

237
 NOTE S

Part I: Introduction

Chapter 1

1. These definitions are analytical. They focus not on nominal occupational roles but on the
characteristics of the particular work situations. Thus not every street-level bureaucrat works for
a street-level bureaucracy [for example, a relocation specialist (a type of street-level bureaucrat)
may work for an urban renewal agency whose employees are mostly planners, builders, and
other technicians]. Conversely, not all employees of street-level bureaucracies are street-level
bureaucrats (for example, file clerks in a welfare department or police officers on routine clerical
assignments).
The conception of street-level bureaucracy was originally proposed in a paper prepared for
the Annual Meeting of the American Political Science Association in 1g6g, "Toward a Theory of
Street-Level Bureaucracy." It was later revised and published in Willis Hawley and Michael
Lipsky, eds., Theoretical Perspectives on Urban Politics (Englewood Cliffs, N.J.: Prentice-Hall,
1977). pp. 1g6-213
2. U.S. Bureau of the Census, Public Employment in 1973, Series GE 73 No. 1 (Washing-
ton, D.C.: Government Printing Office, 1974), p. 9 Presented in Alan Baker and Barbara
Grouby, "Employmen t and Payrolls of State and Local Government s, By Function: October
1973," Municipal Year Book, 1975 (Washington, D.C.: International City Managers Associa-
tion, 1975), pp. 1og--112, table 4/3. Also, Marianne Stein Kah, "City Employmen t and Payrolls:
1975," Municipal Year Book, 1977 (Washington, D.C.: International City Managers Associa-
tion, 1977), pp. 173-179. These figures have been adjusted to represent full-time equivalents.
For purposes of assessing public commitment s to providing services, full-time equivalents are
more appropriate statistics than total employment figures, which count many part-time em-
ployees.
3 Jeffry H. Galper, The Politics of Social Services (Englewood Cliffs, N.J.: Prentice-Hall,
1975), P 56.
4 Lois Forer, Death of the Law (New York: McKay, 1975) p. 191.
5 New York Times, April4, 1976, p. 22.
6. Baker and Grouby, "Employmen t and Payrolls of State and Local Governments.
7 New York Times, July 10, 1977, p. F13.
8. Of four cities with populations over one million responding to a Municipal Year Book
survey, the proportion of personnel expenditures to total expenditures in police departments
averaged 94 percent and did not go below 86 percent. Cities with smaller populations showed
similar tendencies. These observations are derived from David Lewin, "Expenditur e, Compen-
sation, and Employment Data in Police, Fire, and Refuse Collection and Disposal Depart-
ments," Municipal Year Book, 1975 pp. 3!r98, table 1/21. However, the variation was much
greater in the less populous cities because of smaller base figures and the fact that when cities
with smaller bases make capital investments, the ratio of personnel to total expenditures
changes more precipitously.
That public expenditures for street-level bureaucracies go to individuals primarily as salaries
may also be demonstrate d in the case of education. For example, more than 73 percent of all
noncapital education expenditures inside Standard Metropolitan Statistical Areas goes toward
personal services (i.e., salaries). See Government Finances, Number 1, Finances of School Dis-
tricts, 1972 U.S. Census of Government (Bureau of the Census, Social and Economic Statistics
Administration, U.S Department of Commerce), table 4

239
 Notes
9 Many analysts have discussed the increasing role of services in the economy. See Daniel
Bell, The Coming of the Post-Industrial Society: A Venture in Social Forecasting (New York:
Basic Books, 1973); Alan Gartner and Frank Reissman, The Service Society and the Consumer
Vanguard (New York: Harper & Row, 1974); Victor Fuchs, The Service Economy (New York:
Columbia University Press, 1968). On transformations in public welfare, see Gilbert Steiner,
Social Insecurity (Chicago: Rand McNally, 1g66), chap. 1; on public safety, see Allan Silver,
"The Demand for Order in Civil Society," in David Bordua, ed., The Police: Six Sociological
Essays (New York: John Wiley, 1g67), pp. 1-24.
10. Charles Reich, "The New Property, Yale Law journal, vol. 72 (April, 1g64): 733-787.
11. Carl Hosticka, "Legal Services Lawyers Encounter Clients: A Study in Street-Level
Bureaucracy" (Ph.D. diss., Massachusetts Institute of Technology, 1976), pp. 11-13.
12. See Frances Piven' s convincing essay in which she argues that social service workers
were the major beneficiaries of federal programs concerned with cities and poor people in the
1g6os. Piven, "The Urban Crisis: Who Got What and Why," in Richard Cloward and Frances
Piven, The Politics of Turmoil (New York: Vintage Books, 1972) pp. 314-351.
13. J. Joseph Loewen berg and Michael H. Moskow, eds., Collective Bargaining in Govern-
numt (Englewood Cliffs, N.J.: Prentice-Hall, 1972). A. Laurence Chickering, ed., Public Em-
ployee Unions (Lexington, Mass., Lexington Books, 1976); and Margaret Levi, Bureaucratic In-
surgency (Lexington, Mass.: Lexington Books, 1977).
14. The decline is a function of the lower birthrate and periodicity in the size of the school-
age population originally resulting from the birth explosion following World War II. See Baker
and Grouby, Municipal Year Book, 1975, pp. 109ff., on serviceability ratios.
15. This perspective remains applicable in the current period. However, in reaction to this
tendency, programs that would eliminate service mediators and service providers, such as nega-
tive income taxation and housing allowances, have gained support. Fiscal scarcity has brought to
public attention questions concerning the marginal utility of some of these service areas.
16. Consider the New York City policemen who, in October 1976, agreed to work overtime
without pay so that a crop of rookie patrolmen would not be eliminated. New York Times, Octo-
ber 24, 1976, p. 24.
17. There can be no better illustration of the strength of the organized service workers and
their support by relevant interests than the New York State Assembly's overriding of Gov.
Hugh Carey's veto of the so-called Stavisky bill. This legislation, written in a period of massive
concern lor cutting the New York City budget, required the city to spend no less on education
in the three years following the fiscal collapse than in the three years before the crisis, thus tying
the hands of the city's financial managers even more. New York Times, April 4, 1976, p. E6;
April 18, 1976, p. E6.
18. The seminal work here is Robert Rosenthal and Lenore Jacobson, Pygmalion in the
Clossroom (New York: Hoit, Rinehart and Winston, 1968).
19. Martin Rein, "Welfare and Housing," Joint Center Working Papers Series, no. 4 (Cam-
bridge, Mass.: Joint Center for Urban Studies, Spring, 1971, rev. Feb. 1972).
20. On the alleged importance of bureaucratic detachment in processing clients see Peter
Blau, Exchange and Power in Social Life (New York: John Wiley, 1g64), p. 66.
21. See National Advisory Commission on Civil Disorders, Report (New York: Bantam,
1g68); Peter Rossi et al., Roots of Urban Discontent (New York: John Wiley, 1974).
22. Frances Fox Piven and Richard Cloward, Poor People's Movements (New York: Pan-
theon, 1977), pp. 2o-21.
23. Michael Lipsky and Margaret Levi, "Community Organization as a Political Resource,"
in Harlan Hahn, ed., People and Places in Urban Society (Urban Affairs Annual Review, vol. 6)
(Beverly Hills, Calif.: Sage Publications, 1972), pp. 175-199.
24. See James O'Connor's discussion of "legitimation" and his general thesis concerning
the role of the state service sector, in O'Connor, The Fiscal Crisis of the State (New York: St.
Martin's, 1973). On social control functions in particular policy sectors see Samuel Bowles and
Herbert Gintis, Schooling in Capitalist America (New York: Basic Books, 1976); Frances Fox
Piven and Richard Cloward, Regulating the Poor (New York: Pantheon, 1971); Galper, The Poli-
tics of Social Services; Richard Quinney, Criminology (Boston: Little, Brown, 1975); Ira Katznel-
son, "Urban Counterrevolution," in Robert P. Wolff, ed., 1g84 Revisited (New York: Alfred
Knopf, 1973), pp. 139-164.
 Notes

Chapter 2

1. See Chris Argyris, Integrating the Individual and the Organization (New York: John
Wiley, 1g64), pp. 35-41.
2. Frank L. Morris, Sr., "The Advantages and Disadvantages of Black Political Group Ac-
tivity in Two Northern Maximum Security State Prisons" (Ph. D. diss., Massachusetts Institute
ofTechnology, 19]6), p. 40.
3 For some analysts the defining characteristic of professionalism is simply the discretion
to make decisions about clients. In this view street-level bureaucrats would unquestionably be
professionals. See Albert Reiss, The Police and the Public (New Haven: Yale University Press,
1971), p. 122.
4 On rules and the police, see James Q. Wilson, Varieties of Police Behavior (Cambridge:
Harvard University Press, 1g68), p. 31; David Perry, Police in the Metropolis (Columbus, Ohio:
Charles Merrill, 1975), p. 168. See also Gresham Sykes' discussion of the dilemma of prison
guards in being formally required to intervene in all cases of observed infractions, in Sykes, The
Society of Captives (Princeton, N.J.: Princeton University Press, 1958).
5 For example, the Office of Civil Rights of the Department of Health, Education, and
Welfare has responsibility to monitor potential violations as follows: racial discrimination under
Title VI of the Civil Rights Act of 1964 in 16,000 public school districts, 2,8oo institutions of
higher education, and 30,000 institutions of health and social services; in the same areas, dis-
crimination against handicapped people under Section 504 of the Vocational Rehabilitation Act
of 1973; sex discrimination under Section 799A of the Public Health Service Act in 1, 500 health
education institutions, and under Section 745, sex discrimination in nursing schools; sex dis-
crimination under Title IX, Education Amendments of 1972, in 16,000 public school districts;
discrimination by federal contractors under Executive Order u246, innumerable contractors at
863 higher-education campuses, and more than 3,500 additional locations. Virginia Balderama,
"The Office of Civil Rights as a Street-Level Bureaucracy," unpublished seminar paper, Uni-
versity of Washington, March, 19]6.
6. David Perry and Paula Sornoff report that welfare workers' behavior with clients in Cali-
fornia is ruled by ll5 pounds of regulations; that the average police officer is obliged to enforce
approximately 30,000 federal, state, and local laws. Perry and Sornolf, "Street Level Adminis-
tration and the Law: The Problem of Police Community Relations," Criminal Law Bulletin, vol.
8, no. 1 (January-February, 1972), p. 46.
7 Consider police assertions that they would be less willing to risk intervention if civilian
review boards could penalize them for errors of judgment made under hectic and confusing cir-
cumstances which civilians might not appreciate.
8. For a discussion of attempts to introduce uniform sentencing for juvenile offenders, see
the report of the findings of the Juvenile Justice Standards Project, New York Times, November
30, 1975, p. 1; for adult offenders, New York TimBs, October 16, 1977, p. 1.
9 See James Q. Wilson, "The Bureaucracy Problem," The Public Interest (Winter, 1g67),
pp. 3-9
10. Keith Stevenson and Thomas Willemain, "Analyzing the Process of Screening Calls for
Emergency Service" (Cambridge, Mass.: Operations Research Center, Massachusetts Institute
of Technology, September, 1974), Technical Report TR-oB-74.
u. Interviews with administrative personnel, Veterans Administration hospital, BedfOrd,
Mass., August, 1974.
12. Fred Hechinger, "Where Have All the Innovations Gone?" New York Times, No-
vember 16, 1975, p. ED3o.
13. The emphasis here is on structural explanations. Lower-level participants may also per-
sonally disagree with policy objectives. See Donald Van Meter and Carl Van Horn, ''The Policy
Implementation Process: A Conceptual Framework," Administration and Society, vol. 6, no. 4
(1975). pp. 482-483.
14. For this analysis I have drawn on Daluendorf' s observation that assuming ubiquitous
conftict among social units helps in understanding some political events better than assuming
inclinatiq!is toward stability, integration, and interdependence. For Daluendorf, conftict rela-
tions are inevitable since authority relations, which are present in all social units, are necessarily
Notes
relations of subordination and superordination. However, Dahrendorf for general purposes is
unable to choose between the two models of social dynamics-the integration or the "coercion"
model-although for an analysis of class formation and development he favors the coercion
perspective. Similarly, the model stressing conflict outlined here may be applicable under the
circumstances outlined here, while the systems-integration model may be appropriate for other
aspects of policy analysis. Generally see Ralf Dahrendorf, Class and Class Conflict in Industrial
Society (Stanford, Calif.: Stanford University Press, 1959), chap. 5
15. The perspective developed in these paragraphs is elaborated in Michael Lipsky,
"Standing the Study of Public Policy Implementation on Its Head," in W. Dean Burnham and
Martha Wagner Weinberg, eds., American Politics and Public Policy (Cambridge, Mass.: Mas-
sachusetts Institute of Technology Press, 1978), pp. 391-402.
16. Argyris, Integrating the Individual and the Organization, pp. 5g-67.
17. On low motivation of public service workers, see Eric Nordlinger, Decentralizing the
City (Cambridge, Mass.: Massachusetts Institute ofTechnology Press, 1972), chap. 3; and E. S.
Savas and Sigmund Ginsburg, "The Civil Service-A Meritless System?" The Public Interest,
no. 32 (Summer, 1973), pp. 7o--85.
The problem of maintaining worker participation in organizations is a classic issue of organi-
zational theory. For a significant early analysis, see James March and Herbert Simon, Organiza-
tions (New York: John Wiley, 1958).
18. Dahrendorf, Class and Class Conflict in Industrial Society, p. 178.
19. Donald Van Meter and Carl Van Horn point out that the "disposition of implementors"
is critical to policy implementation success. The following discussion elaborates two of the con-
ditions under which, they assert, policy implementors will resist implementation: when the
policies to be implemented offend their sense of self-interest, and when the policies threaten
features of the organization and procedures they desire to maintain. Van Meter and Van Horn,
"The Policy Implementation Process: A Conceptual Framework," pp. 482-483.
20. The discussion is necessarily schematic to a degree. For example, it is an over-
simplification to treat street-level bureaucracies as comprised of lower-level workers and man-
agers. In this discussion the term "manager" refers to someone in an immediate supervisory
position vis-a-vis street-level bureaucrats (for example, a supervisor in a public welfare agency,
a police captain in charge of a precinct sector, or a principal in a nondepartmentalized public
school). "Objectives" refers to the goals that the supervisor is charged with realizing. It is neces-
sary to put it this way because the role of supervisor is itself subordinate to other roles in a
complex bureaucracy. The focus on the divergence of objectives between the organization and
the lowest-level workers could with some modifications be applied to the relations between the
lowest-level supervisor and the roles to which this position is subordinate.
21. Argyris, Integrating the Individual and the Organization, p. 36.
22. For an extended treatment of the sources of street-level bureaucrats' influence, see
Jeffrey Prottas, People-Processing: The Street-Level Bureaucrat in Public Service Bureaucracies
(Lexington, Mass.: Lexington Books, 1979).
23. This paragraph is based upon personal observations, conversations with court person-
nel, and sustained discussions with workers in the Boston Court Resources Project.
24. See Jeffrey Prottas, People-Processing, chapter 3
25. See Jon Pynoos, "Breaking the Rules: The Failure to Select and Assign Public Housing
Tenants Equitably," (Ph.D. diss., Harvard University, 1974).
26. Internal memo, "Budget Bureau Recommendations for Saving in the Welfare Budget,"
March 24, 1g6g, p. IV--6 (author's files).
27. Kuh said he acted under a general provision of the law permitting prosecutors to use
discretion to assure humane and rational dispositions. See New York Times, June 19, 1974, p. 1.
At this date 87 methadone "sellers" were affected by his decision.
28. Pietro Nivola, "Municipal Agency: A Study of the Housing Inspectional Service in Bos-
ton," (Ph.D. diss., Harvard University, 1976, chap. 7).
29. David Mechanic, "Sources of Power of Lower Participants in Complex Or~~;anizations,"
Administrative Science Quarterly, vol. 7, no. 2 (December, 1g6z), pp. 349-364.
30 Ibid., p. 352
31. See Jonathan Rubinstein, City Police (New York: Farrar, Strauss, Giroux, 1973), chap.
2.
 Notes
32. Thomas Scheff, "Control over Policy by Attendants in a Mental Hospital," journal of
Health and Human Behavior, vol. 2(1glh), p. 97, cited in Mechanic, "Sources ofPower," p. 363.

Part II: Conditions of Work

Introduction

1. For vivid descriptions of street-level bureaucracies from the client perspective, see Paul
Jacobs, Prelude to Riot (New York: Vintage, 1g68); Joseph Lyford, The Airtight Cage (New York:
Harper & Row, 1g66).
2. Two recent studies of policy making focusing on the importance of the context of decision
making, particularly lack of resources and uncertainty, are Martha Wagner Weinberg, Manag-
ing the State (Cambridge, Mass.: Massachusetts Institute ofTechnology Press, 1977); Douglas
Yates, The Ungovernable City (Cambridge, Mass.: Massachusetts Institute ofTechnology Press,
1977).
3 In this discussion the word "client" is used to refer to the subjects of interactions of
street-level bureaucrats. This creates several problems in comparison with common usage. For
example, the word "client" commonly is used to refer to people for whom service is performed.
In this sense the clients of police are the people (or general public) being protected, rather than
the subjects of interaction, which include robbers as well as robbed. Also, there are generic
words for the subjects of service which make "client" appear awkward (e.g., doctors' patients,
teachers' students or pupils). However, considering all the difficulties it seems less pedantic to
refer to "clients" rather than "subjects" and truer to the synthetic objectives of this study to
refer often to "clients" rather than to utilize in general discussion the generic words for sub-
jects of study. I trust the reader will bear with me on this point.
For a discussion of the implications of various terms designating the lowest levels of organiza-
tional participation see Amitai Etzioni, A Comparative Analysis ofComplex Organizations (New
York: The Free Press, 1g61), pp. 17-21.

Chapter 3

1. Anthony Downs, Inside Bureaucracy (Boston: Little, Brown, 1g67), p. 3

2. Sheldon Messinger, "Organizational Transformation: A Case Study of a Declining Social
Movement," American Sociological Review, vol. 20, no. 2 (1955), pp. 3-10.
3 Carl Hosticka, "Legal Services Lawyers Encounter Clients: A Study in Street-Level
Bureaucracy," Unpublished (Ph.D. diss., Massachusetts Institute of Technology, 1976).
4 Don Zimmerman, "The Practical Basis of Work Activities in a Public Assistance Organi-
zation," in Donald Hansen, ed., Explorations in Sociology and Counseling (New York:
Houghton Mifllin, 1g6g), pp. 245-249, cited in Jeffrey Prottas, People-Processing: The Street-
Level Btlt'8aucrtJt in Public Service Bureaucracies (Lexington, Mass.: Lexington Books, 1979),
p. 17.
5 Typical case loads per judge per year in selected cities are: Minneapolis District Court,
700; Pittsburgh Common Pleas Court, 1,263; Chicago Preliminary Hearing Court, 2,666 to
7,000. See Martin Levin, "Delay in Five Criminal Courts," journal of Legal Studies, vol. 4, no.
1 (January, 1975), table I, p. 88. Note these figures are for felony cases.

243
Notes
6. See John H. McNamara, "Uncertainties in Police Work: The Relevance of Police Re-
cruits' Backgrounds and Training," in David Bordua, The Police: Six Sociological Essays (New
York: John Wiley, 1g67), pp. 168-177-
7 Maureen Mileski, "Courtroom Encounters: An Observation Study of a Lower Criminal
Court," Law and Society Review, vol. 5, no. 5 (May, 1971), p. 479
8. See Zimmerman, "The Practical Basis of Work Activities in a Public Assistance Organi-
zation."
g. Richard Weatherley points out that paperwork also protects workers from clients and
provides solace from job pressures. Many workers appreciate the required interruptions from
seeing clients, and thus many actually depend upon paperwork routines as job-coping devices
that help moderate the work day.
10. For a good discussion of the transition from rookie to veteran, see John Van Maanen,
"Working the Street: A Developmental View of Police Behavior," in Herbert Jacob, ed., The
Potential for Reform of Criminal justice (Beverly Hills, Calif.: Sage, 1974).
11. See James Q. Wilson, Thinking About Crime (New York: Basic Books, 1975), chaps. 3.
8.
12. See Wilson, Varieties of Police Behavior (Cambridge, Mass.: Harvard University Press,
1g68), pp. 1g-2o.
13. Carl Werth man and Irving Piliavin, "Gang Members and the Police," in Bordua, ed.,
The Police: Six Sociological Essays, p. 74
14. See, for example, William A. Westley, "Violence and the Police," American journal of
Sociology, vol. 59 (August, 1953), p. 39; Werthman and Piliavin, "Gang Members and the
Police," p. 93; Richard Blum, "The Problems of Being a Police Officer," Police (January, 1961)
p. 12.
15. Gresham Sykes, The Society of Captives (Princeton, N.J.: Princeton University Press,
1958).
16. Georgette Bennett-Sandier and Earl Ubell, "Time Bomb in Blue," New York, March
21, 1977. p. 47
17. Alfred M. Bloch, "The Battered Teacher-A New Form of Combat Neurosis," un-
published paper dated March 27, 1976.
18. Howard Becker, "Social Class and Teacher-Pupil Relationships," in Blaine Mercer and
Edwin Carr, eds., Education and the Social Order (New York: Holt, Rinehart and Winston
1957), pp. 278-z7g; Bernard Kelner, How to Teach in Elementary School (New York: McGraw-
Hill, 1958), p. 19.
19. B. L. Margolis et al., "Job Stress: An Unlisted Occupational Hazard," journal of Oc-
cupational Medicine, vol. 16, no. 10 (Oct., 1974), pp. 65g-661. Interestingly, the most consis-
tent relationships between mental health and working conditions occurred with respect to two
other indicators: underutilization of workers' abilities and nonparticipation in decisions affect-
ing one's job.
20. George Kirkham, "What a Professor Learned When He Became a Cop," U.S. News
and World Report, April 22, 1974. p. 72.
21. In the early 1970s the number of emergency room visits incr~ased at the rate of about
10 percent per year. Thomas Willemain, "The Status of Performance Measures for Emergency
Medical Services," Technical Report No. o6-74 (Cambridge, Mass: Operations Research Cen-
ter, Massachusetts Institute ofTechnology, July, 1974), p. 3 This figure is clearly too large to be
accounted for by increases in population or the absolute number of emergencies experienced.
According to a study of Chicago emergency facilities the most important factor in accounting for
an increase in utilization of 82 percent per 1,000 population from 1g6o to 1g6g was the increase
in the number of people using the emergency rooms for treatment of nonemergency conditions.
Barry Schwartz, Queuing and Waiting (Chicago: University of Chicago Press, 1975), p. 127.
22. Catherine Kohler Riessman, "The Supply-Demand Dilemma in Community Mental
Health Centers," American journal of Orthopsychiatry, vol. 40, no. 5 (October, 1970), pp.
8s8-86s
23. This is the hypothesis of Richard A. Posner, "An Economic Approach to Legal Proce-
dure and Judicial Administration," journal of Legal Studies, vol. 2 (1973), pp. 447-448. Cf.
Levin, "Delay in Five Criminal Courts," pp. 127-128.
24. Robert Perlman, Consumers and Social Services (New York: John Wiley, 1975), p. 70.
2!'). For discussions of different aspects of changing expectations of the police see Wilson,

244
 Notes
Varieties of Police Behavior; Arthur Waskow, From Race Riot to Sit-in (Garden City, N.Y.:
Doubleday, 1g66); James Richardson, "To Control the City: The New York Police in Historical
Perspective," in Kenneth T. Jackson and Stanley Schultz, eds., Cities in American History (New
York, Knopf, 1972), pp. 2&r-288; Allan Silver, "The Demand for Order in Civil Society: A
Review of Some Themes in the History of Urban Crime Police and Riot," in David Bordua, ed.,
The Police, pp. 1-24; Robert Fogelson, Big City Police (Cambridge, Mass: Harvard University
Press, 1977).
26. C. H. Goodrich et. a!., "The New York Hospital-Cornell Medical Center: Progress
Report on an Experiment in Welfare Medical Care," American Journal of Public Health, vol.
55, no. 1 (1g65), pp. 88-g3; James Weiss and Merwyn Greenlick, "Determinants of Medical
Care Utilization: The Effect of Social Class and Distance on Contacts with the Medical Care
System," Medical Care, vol. 9 (1970). Cited in Deborah Stone, Institute of Policy Sciences,
Duke University, "Professionals and Social Services," unpublished paper (March, 1976).
27. Carol Ruth Silver, "The Imminent Failure of Legal Services fur the Poor: Why and
How to Limit Caseload," Journal of Urban Law, vol. 46 (1g6g), p. 217.
28. For a similar analysis applied to collective demands see Michael Lipsky and David J.
Olson, Commission Politics: The Processing of Racial Crisis in America (New Brunswick, N.J.:
Transaction Books, 1977), pp. 3-6.
29. Nordlinger makes a well-reasoned argument that out of eighty thousand complaints
concerning city services registered in Boston in 1970, fully fifty to sixty thousand would not have
been registered in the absence of the Little City Halls program. He estimates that at least half of
these calls, representing approximately a third of these service complaints, were legitimate and
not crank complaints or trivial. In other words, at least a third of all the service complaints
received were new, yet of the kind the city had been receiving under old demand-receiving
policies. See Eric Nordlinger Decentralizing the City: A Study of Boston's Little City Halls
(Cambridge, Mass.: Massachusetts Institute ofTechnology Press, 19722, p. 286.
30. Reissman, fur example, cites the case of an Oklahoma public health clinic that increased
tenfold the number of persons brought to the clinic for immuniZation by relying for outreach on
seven paraprofessionals rather than three nurses. Reissman, "The Demand-Supply Dilemma,"
p. 858.
31. David Kirp, "Schools as Sorters: The Constitutional and Policy Implications of Student
Classification," University of Pennsylvania Law Review, vol. 121, no. 4 (April, 1973), p. 712.
32. Some of the inactive caseload may consist simply of cases that would not be on the rolls
if the worker had had time to find out that they should be eliminated fur reasons of changed cir-
cumstances. This enrages agency administrators, particularly in welfare or other entitlement
programs. For a discussion of these caseload dynamics in legal services see Hosticka, "Legal
Services Lawyers Encounter Clients."
33 Here is a concrete if hypothetical illustration. If, on the average, legal service lawyers
have formal case loads of eighty and active case loads of twenty, if another lawyer were added to
a five-person office, and if no new cases were accepted, then each attorney would have approxi-
mately 66 cases (the old case load of the office now divided by six). But each attorney would still
have an active case load of twenty. This increases the number of clients actively served, but
does not improve the situation qualitatively. The pressure of time remains unchanged since the
active case load is a function not of work assigned but of the amount of case pressure that work-
ers can accommodate.
34 Generally, see chap. 11.
35 For example, Milton Heumann has fOund that a loss of adversary activity in the courts
and an increase in plea bargaining apparently are not directly related to case-load pressures. See
Heumann, "A Note on Plea Bargaining and Case Pressure," Law and Society Review, vol. 9, no.
3 (Spring, 1975), PP 515-528.
36. Ibid., p. 527
37 See the discussion in Robert Alford, Health Care Politics (Chicago: University of Chi-
cago Press, 1975), p. 222.
38. For a discussion of costs imposed on clients, see chapter 8 herein.
39 See Downs, Inside Bureaucracy, p. 188.

245
Notes

Chapter 4

1. On goal conflicts in police work see James Q. Wilson, Varieties of Police Behavior
(Cambridge, Mass.: Harvard University Press, 1g68); in public education, see Jeffrey Raffel,
"Responsiveness in Urban Schools: A Study of Adaptation to Parental Preferences in an Urban
Environment," (Ph.D. diss., Massachusetts Institute of Technology, 1972); in public welfare,
see Gilbert Steiner, The State of Welfare (Washington, D.C.: Brookings, 1971).
2: Willis Hawley, "Dealing with Organizational Rigidity in the Public Schools," (paper pre-
sented at the Annual Meeting of the American Political Science Association, September, 1971),
p. 22, n. 77 See also Yeheskel Hasenfeld and Richard English, eds., Human Service Organiza-
tions (Ann Arbor, Mich.: University of Michigan Press, 1974), pp. g-12.
3 Martin Landau, "On the Concept of a Self-Correcting Organization," Public Administra-
tion Review, vol. 33, no. 6 (November-December, 1973), p. 536.
4 Daniel P. Moynihan, Maximum Feasible Misunderstanding (New York: The Free Press,
1g6g), chap. 5
5 Martin Rein, Social Policy (New York: Random House, 1970), p. xi.
6. Hasenfeld and English, Human Service Organizations, pp. 12-14.
7 See Gary Bellow and Jeanne Kettleson, "From Ethics to Politics: Confronting Scarcity
and Fairness in Public Interest Practice," Boston University Law Review, vol. 58, no. 3 (May,
1978), pp. 337-390
8. Amitai Etzioni, A Comparative Analysis of Complex Organizations (Glencoe, Ill.: Free
Press, 1g61),
9 Because teachers, social workers, nurses, and other occupational groups typically display
professional characteristics such as length of training, degree of autonomy, etc., to a lesser
degree than the professions of medicine and law, some analysts have chosen to call them "semi-
professions" to highlight this distinction. See Amitai Etzioni, ed., The Semi-Professions and
their Organization (New York: The Free Press, 1g6g).
10. Martin Levin, "Delay in Five Criminal Courts," journal of Legal Studies, vol. 4, no. 1
(January, 1975), P go.
11. On the tendency for organizations to maintain themselves and enhance their position
see, notably, Chester Barnard, The Functions of the Executive (Cambridge, Mass.: Harvard
University Press, 1938).
12. See then President Gerald Ford's statement proposing to reduce the runaway risks of the
food-stamp program, New York Times, October 10, 1975.
!3. Michael Lipsky and Morris Lounds, "Citizen Participation in Health Care: Dilemmas
of Government Induced Participation," journal of Health Politics, Policy and Law, vol. 1, no. 1
(Spring, 1976), pp. 85-111.
14. Theodore Sarbin and Vernon Allen, "Role Theory," in Gardner Lindzey and Elliot
Aronson, eds., The Handbook of Social Psychology, 2d ed. (Reading, Mass.: Addison-Wesley,
lg68), PP 488-s67. esp. PP 4g8-499, 532.
15. See Wilson, Varieties of Police Behavior; Raffel, "Responsiveness in Urban Schools";
and Steiner, The State of Welfare.
16. On differences in cities' political cultures see Robert Alford, Bureaucracy and Partici-
pation: Political Culture in Four Wisconsin Cities (Chicago: Rang-McNally, 1g6g), and Herbert
Jacob, Debtors in Court: The Consumption of Government Services (Chicago: Rand McNally,
1g6g).
17. Wilson discusses the "zone of indifference" in which police administrators are free to
act in Varieties of Police Behavior, p. 233. The phrase is from Barnard, The Functions of the Ex-
ecutive, p. 167.
18. A case in point was provided by Boston policemen charged with preventing white Bos-
ton residents from harassing black school children when Boston schools were integrated in
1975. The policemen often grew up in the same neighborhoods as those in the crowds they were
trying to control. They were friendly with or lived among neighborhood residents and did not
personally approve of school integration as it was being carried out. See John Kifner, "Men in
the Middle," New York Times Magazine (September 12, 1976), pp. 36ff.
19. The theme of role conflict pervades the literature on police. See, for example, Carl
 Notes
Werthman and Irving Piliavin, "Gang Members and the Police," Albert Reiss and David Bor-
dua, "Environment and Organization: A Perspective on the Police," and James Q. Wilson,
"Police Morale, Reform, and Citizen Respect: The Chicago Case," in Bordua, ed., The Police;
Herman Goldstein. "Police Discretion: The Ideal Versus the Real." Public Administration Re-
view, vol. 23 (September, 19&.3), p. 142; Arthur Niederhoffer, Behind the Blue Shield (New
York: Doubleday, 1g67).
20. Perhaps the extreme expression of the exclusion of clients as a reference group is found
in the courts. As Martin Levin writes, the person "with perhaps the most potential interest in
the criminal court-the victim-usually does not even watch its proceedings, and when he
does, he does not exercise effective supervision. Indeed, almost all aspects of the court .vrocess
. . . operate to discourage his effort to watch." Levin, "Delay in Five Criminal Courts, p. 95
2.1. See Norman Fainstein and Susan Fainstein, Urban Political Movements (Englewood
Cliffs, N.J.: Prentice-Hall, 1974).
22. Research findings in this area are summarized in Sarbin and Allen, "Role Theory," pp.
so:rso6. In their study of job stress Margolis et al. found role ambiguity to be associated with
six out of ten measures of job stress. "Job Stress: An Unlisted Occupational Hazard," pp.
65g-661.
2.3. Downs, Inside Bureaucracy, chap. 3
2.4 For further discussion of problems in measuring performance, see chapter 11. Here the
objective is simply to elaborate the proposition that unavailability of performance measures is
a common, critical condition of street-level bureaucracy work.
2.5. Rubinstein, City Police, pp. 32.-43, 67.
26. John I. Kits use and Aaron V. Cicourel, "A Note on the Use of Official Statistics," Social
Problems, vol. 11 (19&.3), pp. 131-139.
2.7. Peter Blau, The Dynamics of Bureaucracy, rev. ed. (Chicago: University of Chicago
Press, 1g63), chap. 3
z8. See Stanton Wheeler, "The Structure of Formally Organized Socialization Settings," in
'Orville Brim, Jr. and Stanton Wheeler, Socialization after Childhood: Two Essays (New York:
John Wiley, 1g66), pp. 102.ff.
We may hypothesize that the willingness of street-level bureaucrats to develop, and the
public to accept, these surrogate measures of performance reinfOrces conservative tendencies.
When these reified qualities are accepted as good or significant, or even if they simply deter-
mine the reward structure of the agency, the people who display these characteristics remain
entrenched. Significantly, it is highly upsetting to the status quo when these surrogate mea-
sures are challenged. This is why reform in police departments may be enhanced by insisting
that a college degree be a condition of employment. Those who prospered under the old system
are disadvantaged by this innovation. But this would only be true in departments that pre-
viously had a relatively uneducated staff. To take the point to the extreme, it would also be a
destabilizing and possibly advantageous reform in a highly educated department to forbid
employment eX college graduates. Similarly, it might disrupt the status quo to reward teachers
who have previously had outside work experience or to reward teachers who show particular
abilities in interacting with a wide range of students, since these are qualities that are not nor-
mally rewarded by public school systems.
2.9. Hawley, "Organizational Rigidity in the Public Schools," p. 13.
30. James G. Anderson, "The Authority Structure of the School: System of Social
Exchange," Educational Administration Quarterly, vol. 3 (Spring, 1!J67), p. 136, cited in Haw-
ley, "Organizational Rigidity in the Public Schools," p. 13.
31. David Seidman and Michael Couzens, "Crime, Crime Statistics, and the Great Ameri-
can Anti-Crime Crusade: Police Misreporting of Crime and Political Pressures" (Paper pre-
sented at the Annual Meeting of the American Political Science Association, Washington, D.C.
1972.).
32. Genel'l!lly see Pietro Nivola, "Municipal Agency: A Study of the Housing Inspection
Service in Boston." (Ph.D. diss., Harvard University, 1976).
33 Albert 0. Hirschman, Exit, Voice and Loyalty (Cambridge, Mass.: Harvard University
Press, 1970).
34 See Hawley, "Organizational Rigidity in the Public Schools," p. 12..
35 Gary Bridges, "Citizen Choice in Public Services: Voucher Systems," in E. S. Savas,
ed., Alternatives for Delivering Public Services (Boulder, Colo.: Westview Press, 1977), pp.

247
Notes
51-109; David K. Cohen and Eleanor Farrar, "Power to the Parents?-The Story of Education
Vouchers," Public Interest, no. 48 (Summer, 1977), pp. 72-97.

Chapter 5

1. Others have commented on the extent to which clients of some organizations are non vo-
luntary, and they have attempted to assess the importance of this distinction for treatment and
organizational behavior. See Elaine Cumming, Systems of Social Regulation (New York: Ather-
ton, 1g68). James D. Thompson discusses the extent to which two variables (the extent to which
a member of an organization is tightly or loosely controlled by its rules and assumptions and
whether the nonmembers, in our case clients, participate in the interaction voluntarily or not)
affect the relationship between members and nonmembers of an organization. Thompson's
dichotomization of the interaction variable presents the choices in extreme form and does not ac-
commodate gradations. Thompson's two variables when combined yield four organizational types;
our analysis draws attention to the probability that, analytically, the participation of most poor
people in transactions with public agencies tends to be mandatory. See Thompson, "Organiza-
tions and Output Transactions, in Elihu Katz and Brenda Danet, eds., Bureaucracy and the
Public (New York: Basic Books, 1972), pp. 191-211.
2. The extent to which clients can affect doctors' behavior is the subject of Eliot Freidson's
article, "Client and Medical Practice," American journal of SociokJgy, vol. 65 (January, 1g6o),
374-382. See also, Amitai Etzioni, "Administration and the Consumer," American SociokJgical
Quarterly, vol. 3, no. 2 (September, 1955), pp. 257-264.
3 See Willis Hawley, "Organizational Rigidity in the Public Schools," (paper presented at
the Annual Meeting of the American Political Science Association, September, 1971), p. 15.
4 I have found the most accessible discussion of bargaining to be Thomas Schelling, The
Strategy of Conflict (Cambridge, Mass.: Harvard University Press, 1g6o).
5 Julius Roth, "Some Contingencies ofthe Moral Evaluation and Control of Clientele: The
Case of the Hospital Emergency Service," in Yeheskel Hasenfeld and Richard English, eds.,
Human Seroice Organizations, (Ann Arbor, Mich.: University of Michigan Press, 1974), p. 502.
For a general orientation to the perspective implicit here see Erving Gotfman, Strategic In-
teractions (Philadelphia, Pa.: University of Pennsylvania Press, 1g6g).
6. Eliot Freidson, Professional Dominance (New York: Atherton, 1970).
7 See Gresham Sykes, Society of Captives (Princeton, N.J.: Princeton University Press,
1958), pp. 4S-s8.
8. For a discussion of ways inmates are induced to contribute to their own control in mental
hospitals and other institutions that totally circumscribe people's lives see Erving Gotfman,
Asylums (Garden City, N.Y.: Doubleday, 1g61), pp. 177-207.
9 Gotfman, Strategic Interactions.
10. Ira Katznelson, Black Men, White Cities (New York: Oxford University Press, 1973), p.
25. On the conceptualization of research on relationships of dominance and subordination, see
chap. 2.
11. The title of this section is obviously a paraphrase (apt, I trust) of Peter Berger and
Thomas Luckmann, The Social Construction of Reality (Garden City, N.Y.: Doubleday, 1976).
For a detailed elaboration of the process of client categorization, see Jeffrey Prottas, People-
Processing: The Street-Level Bureaucrat in Public Service Bureaucracies (Lexington, Mass.:
Lexington Books, 1979).
12. For the importance of these distinctions in court processing of juveniles see Robert
Emerson, judging Delinquents (New York: Aldine, 1g6g).
13. On the distinction between structure and behavior see Katznelson, Black Men, White
Cities, chap. 2.
14. Elihu Katz and S. N. Eisenstadt, "Some Sociological Observations on the Response of
 Notes
Israeli Organizations to New Immigrants," in Elihu Katz and Brenda Danet, eds., Bureaucracy
and the Public, p. 79
15. Reported by Carl Hosticka, "Legal Services Lawyers Encounter Clients. A Study in
Street-level Bureaucracy" (Ph.D. diss. Massachusetts Institute of Technology, 1976).
16. Carl Werthman and Irving Piliavin, "Gang Members and the Police," in David Bordua,
ed., The Police: Six Sociological Essays (New York: John Wiley, 1g67), p. 87.
17. Michael Iipsky and Margaret Levi, "Community Organization as a Political Resource,"
in Harlan Hahn, ed., People and Politics in Urban Society (Beverly Hills, Calif.: Sage, 1972),
pp. 195-1g6; Michael Iipsky, Protest in City Politics (Chicago, Ill.: Rand McNally, 1970).
18. In coaching the client street-level bureaucrats are only contributing to a process clients
otherwise engage in: maneuvering to secure what they think will provide the best chance or the
most favorable outcome (getting the best judge, teacher, social worker; phrasing words cor-
rectly; having papers ready, etc.).
19. Jon Pynoos, "Breaking the Rules: The Failure to Select and Assign Public Housing
Tenants Equitably," (Ph.D. diss., Harvard University, 1974). (See chap. 2, pp. 21--22).
20. Pietro Nivola, "Municipal Agency: A Study of the Housing lnspectional Service in Bos-
ton," (Ph.D. diss. Harvard University, 1976), chap. 3
21. Alan Keith-Lucas, Decisions about People in Need (Chapel Hill, N.C.: University of
North Carolina Press, 1957), p. 224.
22. Jerome Skolnick describes how defense attorneys instruct clients fur best results to
express regret and perplexity at their own behavior rather than attempt to explain the behavior
away, although thi is their inclination. Skolnick, "Social Control in the Adversary System,"
]oumalofConflictResolution, vol. 11, no. 1 (1967), pp. 59-&7, in Jerome Skolnick and Richard
Schwartz, eds., Society and the Legal Order (New York: Basic Books, 1970), pp. 414-423, cita-
tion at p. 418.
23. The tendency of bureaucrats to treat some part of the population specially, in opposi-
tion to formal mandates, is explored in Brenda Danet," 'Giving the Underdog a Break': Latent
Particularism Among Customs Officials," in Katz and Danet, Bureaucracy and the Public, pp.
329-337. Danet discusses such tendencies in terms of bureaucrats' "latent particularism." From
an organizational standpoint "latent particularism'' is discussed in terms of "debureaucratiza-
tion" in Elihu Katz and S. N. Eisenstadt, "Some Sociological Observations on the Response of
Israeli Organizations to New Immigrants," in Katz and Danet, pp. 73-88. Coaching the client as
discussed here is one form of "debureaucratization." The term is meant to reflect departure
from an ideal type of bureaucratic universalism. However, the term is somewhat awkwarrl since
it implies a departure from a state of bureaucratic fOrmalism. A bureaucracy which hR never
achieved the universalism of the ideal type in common terms cannot usefully be described as
debureacratized.
24. For a general treatment of this perspective, see Goffinan, Strategic Interactions.
25. Joel Handler and Mary Jane Hollingsworth, The Deserving Poor (Chicago: Markham,
1971). See also Kenneth Clark, Dark Ghetto (New York: Harper & Row, 1g65).
26. Goffinan, Asylums; David Mechanic, Medical Sociology: A Selective View (New York:
The Free Press, 1g68), pp. 11sff.
27. Cf. Murray Edelman's view of the "helping professions" in Political Language: Words
That Succeed and Policies That Fail (New York: Academic Press, 1977), chap. 4
28. Robert Rosenthal and Lenore Jacobson, PygmaUon in the Classroom (New York: Holt,
Rinehart and Winston, 1g68); Ray C. Rist, "Student Social Class and Teacher Expectations: The
Self-Fulfilling Prophecy in Ghetto Education," Harvard Educational Review, vol. 40 (August,
1970), pp. 411-451; see Robert Merton, Social Theory and Social Structure (Glencoe, Ill.: The
Free Press, 1957), chap. 11. Self-fulfilling prophecies that result from assigning clients to ca-
tegories without necessarily being affected by the interaction with street-level bureaucrats are
discussed in chapter 10.
29. Rist, "Student Social Class and Teacher Expectations."
30. Emerson, Jucling Delinquents.
31. Richard Weatherley, Reforming Special Education: Policy Implementation from State
Level to Street Level (Cambridge, Mass.: Massachusetts Institute of Technology Press, 1979).
32 Berger and Luckmann, The Social Construction of Reality, p. 30.
33 Eliot Freidson, Profession of Medicine (New York: Dodd Mead, 1974), pp. u6ff.

249
Notes
34 Jerome Skolnick and Richard Schwartz, ''Two Studies of Legal Stigma," Social Prob-
lema, vol. 10 (Fall, 1g&.), 133-142., cited in Maureen Mileski, "Courtroom Encounters," Law
and Society Review, vol. 5. no. 5 (May, 1971), p. 4g6.
35 Handler and Hollingsworth, The Deserving Poor; Hosticka, "Legal Servies Lawyers
Encounter Clients."

Chapter 6

1. On the role of myth in public policy see Murray Edelman, Political Language (New
York: Academic Press, 1977), esp. chap. 1.
2.. A. Cicourel and J. Kitsuse, The Educational Decision Makers (Indianapolis, Ind.: Bobbs-
Merrill, 19&.3), cited in David Kirp, "Schools as Sorters: The Constitutional and Policy Implica-
tions of School Classification," University of Pennsylvania Law Review, vol. 12.1, no. 4 (April,
1973). p. 711.
3 Deborah Stone has elaborated the dilemma of physicians who are asked to act as both
advocates and overseers of the public purse, in Controlling the Medical Profession (Chicago:
University of Chicago Press, forthcoming).
4 Richard Weatherley, Reforming Special Education: Policy Implementation from State
Level to Street Level (Cambridge, Mass.: Massachusetts Institute of Technology Press, 1979).
5 On the tension between supporting and controlling clients see Elaine Cumming, Sys-
tems of Social Regulation (New York: Atherton, 1g68), pp. 6-g.
6. Jerome Skolnick, "Social Control in the Adversary System," in Jerome Skolnick and
Richard Schwartz, Society and the Legal Order (New York: Basic Books, 1970), pp. 42.1-42.2..
7 Peter Blau, Dynamics of Bureaucracy (Chicago: University of Chicago Press, 1g64).
8. Suggested by Judy Riley, "A Case Study of Street-Level Bureaucracy: Child Protective
Services," unpublished seminar paper, University of Washington, 1976.
9 For a review and discussion of political alienation as a psychological construct see Stanley
Greenberg, "Political Alienation and Political Action," in Willis Hawley and Michael Upsky,
eds., Theoretical Perspectives on Urban Politics (Englewood Cliffs, N.J.: Prentice-Hall, 1976),
pp. 176--183.
10. Here I recognize an assumption that, while the subject of considerable debate, remains
ultimately unresolvable: self-actualization, particularly tendencies toward creativity, coopera-
tion, and personal growth, is inherently a human quality that people strive to express if given
the chance and freed from pursuit of necessities. For a discussion of this orientation in organiza-
tional studies see Chris Argyris, "Some limits of Rational Man Organizational Theory," Public
Administration Review, vol. 33 (June, 1973), pp. 2.53-2.67.
For a summary of how alienation is generally utilized to describe relations of work see Fred-
erick Thayer, An End to Hierarchy! An End to Competition! (New York: Franklin Watts, 1973),
pp. 47-48; also Amitai Etzioni, The Active Society (New York: The Free Press, 1g68), chap. 2.1.
11. Etzioni, The Active Society, pp. 618-62.0.
12.. In these paragraphs I have deliberately chosen language to suggest parallels with analy-
ses of the work of industrial workers.
13. On trends in workers' control see Administr.ation and Society, vol. 7, no. 1 (May, 1975),
an issue devoted to this topic.
14. In response to these work-related problems street-level bureaucracies often do attempt
to control the nature of the clientele. This is treated in chapters 7 through 9
15. Very few studies have concentrated on the dynamics over time of routine treatment of
clients by public agencies. See Alan Keith-Lucas, Decisions about People in Need (Chapel Hill,
N.C.: University of North Carolina Press, 1957) for a study of the treatment of welfare clients in
the south in the 1950s. See Waskow, From Race Riot to Sit-In, for a study of changes in non-
routine police practice over time. For speculative study of the implications of modernization in
police departments see James Q. Wilson, "The Police and the Delinquent in Two Cities," in
 Notes
Stanton Wheeler, ed., Controlling Delinquents (New York: John Wiley, 1g68). For a discussion
of developments in the bureaucratic treatment of Israeli immigrants over time see Katz and
Eisenstadt, 'The Response of Israeli Organizations to New Immigrants," in Elihu Katz and
Brenda Danet, eds., Bureaucracy and the Public (New York: Basic Books, 1972), pp. 73--88.

Part III: Patterns of Practice

Introduction

1. Cf. James Q. Wilson, "The Bureacucracy Problem," Public Interest, no. 6 (Winter,
1g67). pp. 3~
2. For another analysis that assumes people "desire to do a good job," see Downs, Inside
Bureaucracy (Boston: Little Brown, 1g67), p. 1g8.
3 Lee Rainwater, ''The Revolt of the Dirty Workers,"' Transaction, vol. 5. no. 1 (Nov.,
1g(ry-), pp. 2ff.
4 James March and Herbert Simon, Organizations (New York: John Wiley, 1958); Charles
Lindblom, "The Science of'Muddling Through'," Public Administration Review, vol. 19 (Spring
1959). pp. 7g--88.
5 Useful conceptual distinctions for various phenomena related to coping are provided by
Richard Lazarus in Psychological Stress and the Coping Process (New York: McGraw-Hill,
1g66), chap. 1.
6. Routines are the regularized or habitual patterns by which tasks are performed. Simplifi-
cations are symbolic constructs in terms of which decisions about potentially complex phenom-
ena are made, utilizing a smaller set of cues than those presented by the phenomena. Routines
are behavioral patterns of response; simplifications are mental patterns of ordering data with
which routines may or may not be associated. Peter Berger and Thomas Luckmann discuss the
ubiquitous nature of routinization and simplification in The Politics of Everyday Life (Garden
City, N.Y.: Doubleday, 1976), pp. 28ff and 53if. They use the terms "habituations" for the for-
mer, "typifications" for the latter.
7 See Reinhard Bendix, Work and Authority in Industry (New York: John Wiley, 1956),
chap. 4 Willis Hawley presents an interesting discussion of the significance of routinization
in 'The Possibilities of Nonbureaucratic Organization," Willis D. Hawley and David Rogers,
eds., Improving the Quality of Urban Management (Beverly Hills, Cali: Sage, 1974), pp.
371-.p6.
8. See Victor Thompson, Modem Organization (New York: Knopf, 1g61), p. 14; James
March and Herbert Simon, Organizations (New York: John Wiley, 1958), p. 142.
9 See the discussion of"Categorization of Data" in Thompson, Modem Organizations, p.
17
10. The importance of routines in developing policy in other areas is well established. For a
concise general treatment, see Ira Sharkansky, The Routines of Politics (New York: Van Nos-
trand, 1970).
11. On the meaning of the term "political" see David Easton, A Framework for Political
Analysis (Englewood Cliffs, N.J.: Prentice-Hall, 1g65), p. so; and Harold Lasswell, Who Gets
What, When, How? (New York: McGraw Hill, 1936).
12. Karen Orren, Corporate Power and Social Change (Baltimore: Johns Hopkins, 1973).
13. James Davis and Kenneth Dolbeare, Little Groups of Neighbors (Chicago: Markham,
1g68).
14. See Thompson, Modem Organizations, pp. 168--16g.
15. Julius Roth, "Some Contingencies of the Moral Evaluation and Control of Clientele:
The Case of Hospital Emergency Services," in Yeheskel Hasenfeld and Richard English, eds.,
Human Service Organizations (Ann Arbor, Mich.: University of Michigan Press, 1974), p. 500,
italics omitted.
Notes

Chapter 7

1. The reactive nature of police work, and police dependence upon citizens in this respect,
is stressed in Albert Reiss, The Police and the Public (New Haven: Yale University Press,
1971).
2. The latter case is cited by Barry Schwartz, Queuing and Waiting (Chicago: University of
Chicago Press, 1975), p. 24 This excellent volume provides many insights into issues of priori-
ties in client treatment and the costs of seeking service.
3 See Robert Dalll, "The Analysis oflnftuence in Local Communities," in Charles Adrian,
ed., Social Science and Community Action (East Lansing, Mich.: Michigan State University
Press, 1g6o), p. 32
4 See generally Jonathan Rubinstein, "Suspicions," in City Police (New York: Farrar,
Straus, 1973).
5 For example, one prosecutor's office that switched from telephone to mail complaint
handling in processing white collar crimes experienced a 25 percent reduction in complaints
received. Michael Brintnall, "The Allocation of Services in the Local Prosecution of Economic
Crime" (Ph.D. diss., Massachusetts Institute of Technology, 1977), chap. 6.
6. See Schwartz, Queuing and Waiting, chap. 6.
7 When court clerks use confusing legal language we may call it "bureaucratic language as
incantation." Edelman, Political Language: Words that Succeed and Policies that Fail (New
York: Academic Press, 1977), p. g8. But what shall we call the court clerk's chant that strings
words together indistinguishably? Perhaps it should be called "incantation as symbolic lan-
guage." For attempts to deal positively with the rationing effects oflegallanguage, consider the
New York state law requiring consumer contracts to be written in clear, understandable lan-
guage. See New York Times, Aug. 11, 1977, p. B1.
8. This is a paraphrase of the definition of demands in David Easton, A Framework for Po-
litical Analysis (Englewood Cliffs, N.J.: Prentice-Hall, 1g65), p. 120.
9 New York Times, September 25, 1977.
10. Leon Mayhew, "Institutions of Representation: Civil Justice and the Public," Law and
Society Review, vol. 9, no. 3 (Spring, 1975), p. 403. The discrepancy is so great that it would be
difficult to attribute it to differences in the nature of the sample.
11. Richard Cloward and Frances Fox Piven, "A Strategy to End Poverty," The Politics of
Turmoil (New York: Vintage, 1975), pp. 8g--1o6.
12. Gilbert Steiner, The State of Welfare (Washington, D.C.: Brookings, 1971).
13. Kitsuse and Cicourel have written that statistics reflect a great deal about the organiza-
tions collecting the statistics. John Kitsuse and Aaron Cicourel, "A Note on the Uses of Official
Statistics," Social Problems, vol. 11 (1963), pp. 131-139. Sometimes the statistics collectors are
not the same as those formally charged with providing information about services.
14. Eric Nordlinger, Decentralizing the City (Cambridge, Mass.,: Massachusetts Institute
ofTechnology Press, 1972), p. 286.
15. The dynamics of this process are discussed in Michael Lipsky and Morris Lounds,
"Citizen Participation and Health Care: Problems of Government Induced Participation," Jour-
nal of Health Politics, Policy and Law, vol. 1, no. 1 (Spring, 1976), pp. 85-111.
16. See the discussion of the psychological implications of waiting in Schwartz, Queuing
and Waiting, chaps. 1, 8.
17. Virtually every commentary on welfare practices draws attention to the degradation of
clients. See Alan Keith-Lucas, Decisions about People in Need (Chapel Hill, N.C.: University of
North Carolina, 1957); Steiner, The State of Welfare (Washington, D.C.: Brookings, 1971);
Piven and Cloward, Regulating the Poor (New York: Pantheon, 1971), chaps. 4-5.
18. Jeffrey Prottas, People-Processing: The Street-Level Bureaucrat in Public Service Bu-
reaucracies (Lexington, Mass.: Lexington Books, 1979). On the continuing relationships be-
tween ghetto fathers who have deserted and their families, see Elliot Liebow, TaUy's Corner
(Boston: Little, Brown, 1g67).
19. June Grant Wolf, 'The Initial Evaluation at a Walk-In Clinic: Applicant's and Evalua-
tor's Perspectives" (Ph.D. diss., Boston University, 1974), p. 76.
 Notes
20. First-come, first-served, "constitutes the normative basis for most forms of queueing."
Schwartz, Queuing and Waiting, p. 93
21. Ibid., chap. 6
22. Carl Hosticka, "~Services Lawyers Encounter Clients: A Study in Street-Level Bu-
reaucracy" (Ph. D. diss., Massachusetts Institute of Technology, 1976).
23. Catherine Kohler Reissman, "The Supply-Demand Dilemma in Community Mental
Health Centers," American Journal of Orthopsychiatry, vol. 40. no. 5 (October, 1970), p. 86o.
24. See chap. 2.
25. See Jeffry Galper, The Politics of Social Services (Englewood Cliffs, N.J.: Prentice-
Hall, 1975), pp. 70-71.
26. Reissman, "The Supply-Demand Dilemma," p. 86o.
27. See Schwartz, Queuing and Waiting, pp. 2&-29.
28. Ibid., chap. 5 and fu. 5. p. 201.
29. Robert Perlman, Consumers and Social Services (New York: John Wiley, 1975), p. 77
30. Speaker, Annual Convention of the National Legal Aid and Defenders Association,
Seattle, Washington, November, 1975.
31. Nivola, "Municipal Agency: A Study of Housing lnspectional Service in Boston," chap.
3
32 "Budget Bureau Recommendations for Savings in the Welfare Budget," March 24,
1g6g. Unpublished document in author's files.
33 Ibid.
34 Ibid.
35 Ibid.
36. New York Times, December 21, 1977

Chapter 8

1. This is developed in Jeffrey Prottas, People-Processing: The Street-Level Bureaucrat in

Public Seroice Bureaucracies (Lexington, Mass.: Lexington Books, 1979).
2. The term has come into currency to classifY city neighborhoods, arrogantly differentiat-
ing among those that are likely to recover if assisted with government funds, those too far dete-
riorated to save, and those likely to continue to thrive without government assistance.
3 See David Kirp's summary of the evidence in "Schools as Sorters: The Constitutional
and Policy Implications of Student Classification," University of Pennsylvania lAw Review, vol.
121, no. 4 (April, 1973), pp. 705-797.
4 My knowledge of this program was gained over the five-year period in which I served
regularly as a project-site visitor.
5 Brenda Danet, " 'Giving the Underdog a Break': Latent Particularism among Customs
Officials," in Elihu Katz and Brenda Danet, eds., Bureaucracy and the Public (New York: Basic
Books, 1973), pp. 32!r337
6. Nicholas Alex, Black in Blue (New York: Appleton, 196g).
7 Robert Emerson, Judging Delinquents (New York: Aldine, 1g6g).
8. Barney Glaser and Anselm Strauss, "The Social Loss of Dying Patients," American Jour-
nal of Nursing, vol. 64 Gune, 1g64), pp. 111r121.
9 Julius Roth, "Some Contingencies ofthe Moral Evaluation and Control of Clientele: The
Case of the Hospital Emergency Room," in Yeheskel Hasenfeld and Richard English, eds.,
Human Service Organizations, (Ann Arbor, Mich.: University of Michigan Press, 1974), pp.
49!r516.
10. David Sudnow, "Normal Crimes: Sociological Features of the Penal Code in a Public
Defender's Office," Social Problems, vol. 12 no. 3 (Winter, 1g65), pp. 255-276; Emerson,Judg-
ing Delinquents.

253
Notes
11. Robert Scott, "The Selection of Clients by Social Welfare Agencies: The Case of the
Blind," in Hasenfeld and English, Human Seroice Organizations, pp. 485-498; Donald Schon,
'The Blindness System," Public Interest, vol. 18 (Winter, 1970), pp. 25-:38.
12. LDis Forer, Death of the Law (New York: McKay, 1975).
13. Sociologist Jules Henry called the tendency of teachers to concentrate on only a few
students "partial withdrawal." See Henry, "White Peoples Time, Colored Peoples Time,"
Transaction, vol. 2, no. 3 (March-April, 1g65), p. 32
14. See New York Times, March 3, 1977, p. 33
15. Sudnow, "Normal Crimes." See also Erving Goffinan, Relations in Public (New York:
Basic Books, 1971), chap. 6.
16. Maureen Mileski, "Courtroom Encounters," Law and Society Review, vol. 5, no. 5
(May, 1971), p. 513.
17. Carl Hosticka, "Legal Services Lawyers Encounter Clients: A Study in Street-Level
Bureaucracy," (Ph.D. diss., Massachusetts Institute ofTec~nology, 1976).
18. Mileski, "Courtroom Encounters," p. 503.
19. Eliot Freidson, Profession of Medicine (New York: Dodd, Mead, 1970), p. 257
20. Children's Defense Fund, Children Out of School in America (Washington, D.C.:
Children's Defense Fund, 1974).
21. Carl Werthman and Irving Piliavin, "Gang Members and the Police," in David Bordua,
ed., The Police (New York: John Wiley, 1g67), p. 76.
22. Richard Quinney, Criminology (Boston: Little, Brown, 1975), chap. 6; Forer, Death of
the Law, chap. 5.
23. David Mechanic, Medical Sociology (New York: Free Press, 1g68), pp. usff.
24 See chap. 5 herein, note 28.
25. For a convincing discussion of the subjective bases of doctors' views of their work, see
Freidson, Profession of Medicine, pp. 168-172. The subjectivity of teachers' views of ghetto
students is suggested in Peter Rossi et al. The Roots of Urban Discontent (New York: John
Wiley, 1974), P 355
26. For a balanced discussion of the relationship between racial prejudice and generally
biased behavior in judicial sentencing see Willard Gaylin, Partial Justice: A Study of Bias in
Sentencing (New York: Vintage Books, 1975), chap. 3

Chapter g

1. Settings may also be designed to encourage client penetration of these barriers. Deut-
scher describes the case of the public-housing applications officer who was pleased to have the
receptionist's desk to one side so that prospective applicants would come directly to her desk,
bypassing the receptionist. Irwin Deutscher, 'The Gatekeeper in Public Housing," in Irwin
Deutscher and Elizabeth J. Thompson, eds. Among the People: Encounters with the Poor (New
York: Basic Books, 1g68), p. 49
2. See Stanton Wheeler, 'The Structure of Formally Organized Socialization Settings," in
OrviUe Brim, Jr. and Stanton Wheeler, Socialization after Childhood: Two Essays (New York:
John Wiley, 1g66), p. g8.
3 Ibid., p. g8. Wheeler observes that adult socialization agencies implicitly conspire to
present the client process as benign. Sometimes the conspiracy is not so implicit.
4 Street-level bureaucracies must continually justify themselves not only to a client public,
but also to a public constituency concerned with bureaucratic efficiency and effectiveness. To
carry out these tasks street-level bureaucracies expend considerable effort on public relations.
On the public relations budgets of police departments, see LDis Forer, Death of the Law, (New
York: McKay, 1975), p. 176. For an account of police affairs by a New York City deputy commis-
sioner in charge of public relations see Robert Daley, Target Blue: An Insider's View of the
N.Y.P.D. (New York: Delacorte Press, 1973).

254
 Notes
5 Judy Riley, "A Case Study of Street-Level Bureaucracy: Child Protective Services," .un-
published seminar paper, University of Washington, 1976; Pietro Nivola, "Municipal Agency: A
Study of Housing Inspectional Services in Boston" (Ph.D. diss., Harvard University, 1976).
6. Carl Hosticka, "Legal Services Lawyers Encounter Clients: A Study in Street-Level Bu-
reaucracy," (Ph.D. diss., Massachusetts Institute ofTechnology, 1976).
7 Joel Handler and Mary Jane Hollingsworth, The Deseroing Poor (Chicago: Markham,
1971).
8. Howard Becker, "Social Class and Teacher-Pupil Relationships," in Blaine Mercer and
Edwin Carr, eds., Education and the Social Order (New York: Rinehart, 1957), pp. 278-279.
9 Bernard G. Kelner, How to Teach in Elementary School (New York: McGraw-Hill,
1958), p. 19. Also see Willis Hawley, "Dealing with Organizational Rigidity in Public Schools:
A Theoretical Perspective" (paper prepared for delivery at the Annual Convention of the Ameri-
can Political Science Association, September, 1971), p. 6.
10. Arthur Niederholfer, Behind the Blue Shield (New York: Doubleday, 1967), p. 53
11. Jonathan Rubinstein, City Police (New York: Farrar, Straus, 1973), pp. 301-316.
12. John Van Maanen, "Working the Street: A Developmental View of Police Behavior," in
Herbert Jacob, ed., The Potential for Reform of the Criminal Justice System (Beverly Hills,
Calif.: Sage, 1974).
13. See Jerome Skolnick, Justice without Trial (New York: John Wiley, 1g67), pp. 45-46.
See also Rubinstein, City Police, chap. 6.
14. David Kirp makes this point in an unpublished paper, "The Bureaucratization of Child-
hood."
15. Hosticka, "Legal Services Lawyers Encounter Clients."
16. Maureen Mileski, "Courtroom Encounters," Law and Society Review, vol. 5, no. 5
(May, 1971), p. 503.
17. Dennis Trees, unpublished seminar paper, University of Washington, 1976.
18. Mileski, "Courtroom Encounters," pp. 529-530.
19. James D. Thompson, Organizations in Action (New York: McGraw-Hill, 1g67), p. 123.
20. Richard Weatherley, Reforming Special Education: Policy Implementation from State
Level to Street Level (Cambridge, Mass.: Massachusetts Institute of Technology Press, 1979).
21. Nivola, "Municipal Agency."
22. Jeffrey Prottas, People-Processing: The Street-Level Bureaucrat in Public Service
Bureaucracies (Lexington. Mass.: Lexington Books, 1979), chap. 4
23. Hosticka, "Legal Services Lawyers Encounter Clients," provides a persuasive discus-
sion of the influence of receptionists. See generally, David Mechanic, "Sources of Power of
Lower Participants in Complex Organizations," Administrative Science Quarterly, vol. 7, no. 2
(December, 1g62), pp. 349-364.
24. Deutscher, "The Gatekeeper in Public Housing."
25. Joel Handler, "The Juvenile Court and the Adversary System: Problems of Function
and Form," Wisconsin Law Review, vol. 17 (Winter, 1g65).
26. Adoption-agency social workers' "nominal function of advocate for the child and advisor
to the court on the range of alternatives and possible outcomes relative to a child's development
has been transformed into a de facto assumption of judicial powers." Thomas E. Nutt and John
A. Snyder, Trans-Racial Adoption (Study supported under NIMH grant #R03 MH 1g8os-o1),
p. 19
27. Jerome Carlin, "Courts and the Poor" (paper prepared fur delivery at the 1g66 Annual
Meeting of the American Politicai Science Association, New York, September, 1g66), p. 3
28. See, e.g., Mental IUness and Due Process (Ithaca, N.Y.: Cornell University Press,
1g62), cited in Carlin, ibid.
zg. Ray C. Rist, "Student Social Class and Teacher Expectations: The Self-Fulfilling
Prophecy in Ghetto Education," in Yeheskel Hasenfeld and Richard English, eds., Human Ser-
vice Organizations (Ann Arbor, Mich.: University of Michigan Press, 1974), pp. 517-539.
30. David Sudnow, "Normal Crimes: Sociological Features of the Penal Code in a Public
Defender's Office," Social Problems, vol. 12 (Winter), pp. 255~76.
31. Julius Roth, "Some Contingencies of the Moral Evaluation and Control of Clientele:
The Case of the Hospital Emergency Room," in Hasenfeld and English, eds., Human Service
Organizations, pp. 503-504
32. Prottas, People-Processing.

255
Notes
33 An opposite practice, which is also consistent with conserving resources, is the ten-
dency to pass the buck for making determinations to other agencies. This is the current com-
plaint of prison reformers who regard indeterminate sentences, originally intended to permit
prisoners to demonstrate redeeming characteristics, as functioning to keep them under the con-
trol of prison officials who can manipulate the extension or reduction of sentence.
34 Robert Perlman, Consumers and Social Services (New York: John Wiley, 1975), p. 67.
35 Jeffry Galper, The Politics of Social Services (Englewood Cliffs, N.J.: Prentice-Hall,
1975). p. 70.
36. Rikva Bar-Yosefand E. O, Schild, "Pressures and Defenses in Bureaucratic Roles," in
Elihu Katz and Brenda Danet, eds., Bureaucracy and the Public (New York: Basic Books, 1973),
p. ~5
37 For a useful discussion of the function of appeals in the selective service system see
James W. Davis, Jr. and Kenneth Dolbeare, Little Groups of Neighbors (Chicago: Markham,
1g68), chap. 5
38. On the difficulty of filing complaints against the police see Walter Gellhorn, When
Americans Complain (Cambridge, Mass.: Harvard University Press, 1966), pp. 186fT.
39 Frances F. Piven and Richard A. Cloward, Regulating the Poor: The Functions of
Public Welfare (New York: Pantheon, 1971), p. 173. Piven and Cloward attribute the low
number of appeals to the control of the welfare system over clients, resulting in their acquies-
cence to the system of welfare on its terms, a thesis consistent with earlier arguments in this
book.
40. David C. Perry and Paula Sornoff, "Politics at the Street Level: The Select Case of
Police Administration and the Community" (rev. version of a paper presented to the Annual
Meeting of the American Political Science Association, Washington, D.C., 1972), pp. 62~3.
41. Michael Lipsky, Protest in City Politics (Chicago: Rand McNally, 1970), chap. 5
42. See, for example, David C. Perry and Paula Sornoff, "Street Level Administration and
the Law: The Problem of Police-Community Relations," Criminal Law Bulletin, vol. 8, no. 1
Ganuary-February, 1972), p. 54
43 Albert Reiss, Jr., The Police and the Public (New Haven: Yale University Press, 1971),
p. 125
44 The word "emergency" is rarely defined in studies of public services except in the spe-
cific context in which it is applied. See for example, Morris Schwartz and Charlotte Green
Schwartz, Social Approaches to Mental Patient Care (New York: Columbia University Press,
1964), p. so; Egon Bittner, "Police Discretion in Emergency Apprehension of Mentally Ill Per-
sons," Social Problems, vol. 14 (1967), pp. 278-292; Freidson, Profession of Medicine, p. u8.
45 Lipsky, Protest in City Politics, p. Bg.
46. Freidson, Profession of Medicine, p. u8.
47 See Michael Zubkoff, "Emergency Room Service," in Eli Ginsberg, ed., Urban Health
Seroices (New York: Columbia University Press, 1971), pp. ug-124

Chapter 10

1. For a review of the literature on reconciliation of psychological dissonance, see David

Sears and Richard Whitney, Political Persuasion (Morristown, N.J.: General Learning Press,
1973).
2. For a discussion of changes in recruits' attitudes toward their jobs over time see John H.
McNamara, "Uncertainties in Police Work: The Relevance of Police Recruits' Backgrounds and
Training," in David Bordua, ed., The Police: Six Sociological Essays (New York: John Wiley,
1g67), pp. 163-252. See also Richard Cloward and Irwin Epstein, "Private Social Welfare's
Disengagement from the Poor: The Case of Family Adjustment Agencies," in M. Zald, ed.,
Social Welfare Institutions (New York: John Wiley, 1965), pp. 623~43
 Notes
3 Eliot Freidson, Profession of Medicine (New York: Dodd, Mead, 1974), p. Bg.
4 Throughout this study I have focused attention on relationships of street-level bureau-
crats to their work. I do not explore here other causes of worker alienation, which include fac-
tors outside the work situation. For a brief discussion of absenteeism and other employee be-
haviors as adaptive responses to work see Chris Argyris, Integrating the Individual and the
Organization (New York: John Wiley, 1!)64), chap. 4
5 The extent to which public employees are protected from being fired and thus may have
relatively extreme attitudes of withdrawal from engagement in their work, yet still retain their
jobs, is discussed in Eric Nordlinger, Decentralizing the City (Cambridge, Mass: Massachusetts
Institute of Technology Press, 1972), chap. 3
6. To some readers it may seem unnecessary to review the reasons that street-level bureau-
crats continue to work for organizational objectives to some degree. But in many societies with a
less pronounced work ethic, the problem of getting workers to contribute their labor cannot be
taken for granted. See, for example, the analysis of public service employment in Judith Chubb,
"The Organization of Consensus in a Large Southern Italian City: The Social Bases of an Urban
Political Machine" (Ph.D. diss., Massachusetts Institute ofTechnology), 1978.
For a study that analyzes the problem of organizations from the perspective of obtaining
member contributions to organizations, see James March and Herbert Simon, Organizations
(New York: John Wiley, 1958).
7 Aaron Wildavsky, "The Strategic Retreat on Objectives," in Wildavsky, Speaking Truth
to Power: Policy Analysis as a Problem (Boston: Little, Brown, 1979).
8. On orientations of judges toward the poor under various circumstances see Jerome
Carlin, "Courts and the Poor" (paper prepared fur delivery at the 1g66 Annual Meeting of the
American Political Science Association, New York, September, 1g66), p. 7; on teacher orienta-
tions see Howard Becker, "Social Class and Teacher-Pupil Relationships," in Blaine Mercer
and Edwin Carr, eds., Education and the Social Order (New York: Holt, Rinehart and Win-
ston, 1957); on styles of policing, see James Q. Wilson, Varities of Police Behavior (Cam-
bridge, Mass.: Harvard University Press, 1g68).
9 For an effective description of the dilemmas of a public defender in selecting cases for
special attention, see Arthur Rosett and Donald Cressey, Justice by Consent (New York: Lip-
pincott, 1976), chap. 6.
10. On other functions of selective recruitment see Anthony Downs, Inside Bureaucracy
(Boston: Little, Brown, 1g67), pp. 228-233.
11. Harold Wilensky and Charles Lebeaux, Industrial Society and Social Welfare (New
York: Russell Sage, 1958), pp. 233-265.
12. See Richard Weatherley and Michael Lipsky, "Street-Level Bureaucrats and Institu-
tional Innovation: Implementing Special Education Reform," Harvard Educational Review,
vol. 47, no. 2 (May, 1977), pp. 171-197.
13. Victor Thompson has discussed the extent to which ideology serves as a psychological
defense in Modern Organization (New York: Knopf, 1961), pp. 114-137.
14. See Donald Cressey, "Achievement of an Unstated Organizational Goal," in Amitai Et-
zioni, ed., Complex Organizations (New York: Holt, Rinehart and Winston, 1g61), pp. 168-176.
15. See the discussion of overconformity and other defenses in Rivka Bar-Yosef and E. 0.
Schild, "Pressures and Defenses in Bureaucratic Roles," in Elihu Katz and Brenda Danet, eds.,
Bureaucracy and the Public (New York:. Basic Books, 1973), pp. 288-299.
16. See Michel Crozier, The Bureaucratic Phenomenon (Chicago: University of Chicago
Press, 1964), pp. 22o-224.
17. Carl Hosticka, "Legal Services Lawyers Encounter Clients: A Study in Street-level Bu-
reaucracy," (Ph.D. diss. Massachusetts Institute ofTechnology, 1976).
18. Robert Perlman, Consumers and Social Services (New York: John Wiley, 1975).
19. Anticipation of the consequences of their actions appears to be typical of workers on all
levels of the criminal justice system. When New York state passed new drug laws with manda-
tory sentences in 1973, the New York City police department declined to change enfOrcement
strategies because "it was feared that increasing the number of drug arrests . . . would create
intolerable delays in processing cases in the courts." Anthony Japha et. a!., "The Effects of the
1973 Drug Laws on the New York State Courts" (New York, 1976), pp. 2-3.
20. William Ryan, Blaming the Victim (New York: Random House, 1976).

257
Notes
21. Jonathan Kozol, Death at an Early Age (New York: Bantam, 1g67), pp. 10-19.
22. Kenneth Clark has analyzed theories of racial inferiority and cultural deprivation as
functional equivalents in Dark Ghetto (New York: Harper & Row, 1g65), pp. 12sff.
23. See Murray Edelman, Political Language: Words that Succeed and Policies That Fail
(New York: Academic Press, 1977).
24. Erving Coffman, Asylums (Chicago: Aldine, 1g61), pp. 86-87.
25. On differences between whites and blacks regarding ghetto residents' capabilities and
reasons for failures, see Peter Rossi et al., Roots of Urban Discontent (New York: John Wiley,
1974). On some differences in attributions of client responsibility between street-level bureau-
crats, see Clarence Stone, "Paternalism Among Social Agency Employees," Journal of Politics,
vol. 39 (August, 1977), pp. 794-8o4.
z6. Becker, "Social Class and Teacher-Pupil Relationships," pp. 278--299.
27. Judy Riley, "A Case Study of Street-level Bureaucracy: Child Protective Services" (un-
published seminar paper, University of Washington, 1976).

Part IV: The Future of Street-Level Bureaucracy

Chapter 11

1. See Edward Wynne, "Accountable to Whom?" Society, vol. 13 no. z (January/February,

1976), pp. 3<>-37
2. This is not the case with all buffer roles played by people who represent organizations to
the public. For example, salespeople are not expected to be responsible to buyers in the same
sense that, say, social workers are expected to be responsible to clients. See the discussion of
buffer roles in James D. Thompson, "Organizations and Output Transactions," in Elihu Katz
and Brenda Danet, eds., Bureaucracy and the Public (New York: Basic Books, 1973), pp.
191-211.
3 I am not arguing that discretion never can and never should be reduced. On the con-
trary, where lower-level workers usurp discretionary powers it is obviously appropriate for
management to intervene. [For an example of such usurpation see Irwin Deutscher, "The
Gatekeeper in Public Housing," in Deutscher and Elizabeth J. Thompson, eds., Among the
People: Encounters With the Poor (New York: Basic Books, 1g68), pp. 38--sz.] However, when
instances of appropriately circumscribed discretion are exhausted the basic work of street-level
bureaucrats remains.
4 For a discussion of the problems of record keeping and accountability in medicine see
Eliot Freidson, ''The Development of Administrative Accountability in Health Services," Amer-
ican Behavioral Scientist, vol. 1g, no. 3 (January/February, 1976), pp. 286-zgS.
5 The best discussion of the effects of weak management sanctions on developing norms of
reciprocity supportive oflow levels of effectiveness is Eric Nordlinger, Decentralizing the City,
A Study of Boston's Little City HaUs (Cambridge, Mass.: M.I.T. Press, 1972), chap. 3
6. Murray Edelman discusses the symbolic implications of administration and bureaucracy
for mass democracy in The Symbolic Uses of Politics (Urbana, Ill.: University of Illinois Press,
1!)64), chap. 3
7 James Q. Wilson describes this tendency for police departments. ''The police supervisor
. . . would have to judge his patrolmen on the basis of their ability to keep the peace on the
beat, and this . . . is necessarily subjective and dependent on close observations and personal
liuniliarity. Those departments that evaluate officers by 'objective' measures (arrests and traffic
tickets) work against this ideal ... ."Varieties of Police Behavior (Cambridge, Mass.: Harvard
University Press, 1g68), p. 291.
8. Peter Bfau, Tlie DynamicsofBureaucracy, rev. ed. (Chicago: University ofChicago Press,
1!)64), pp. 36--56.
9 James D. Thompson, Organizations in Action (New York: McGraw-Hill, 1g67), p. 123.
 Notes
10. David Seidman and Michael Couzens, "Crime, Crime Statistics, and the Great Ameri-
can Anti-Crime Crusade: Police Misreporting of Crime and Political Pressures" (paper pre-
sented at the Annual Meeting of the American Political Science Association, Washington, D.C.,
1972). Perhaps because they are subject to considerable scrutiny, illustrations of manipulation
of statistics by the police are more likely to come to public attention than other public service
agencies. See, for example, the criticism of an experiment in Orange County, California, that
provided incentive pay increases to police officers for crime reduction. A report on this experi-
ment alluded to the "possibility that the increase in larceny represents a shifting of criminal ac-
tivities or a reclassification of burglaries into a closely related category which will not harm pros-
pects for an incentive reward," New York Times, November 10, 1974. p. 77 Also, New York
Times, May 12, 1972, p. 1.
11. Significantly, the literature on productivity in public service provision draws its most
persuasive examples from these and similar cases of resource deployment, not from the human
services area. See, for example, Edward K. Hamilton, "Productivity: The New York City
Approach," Public Administration Review (November/December, 1972), pp. 784-795.
12. A good discussion of these problems of inference is found in Harry Hatry, "Issues in
Productivity Measurement for Local Governments," Public Administration Review (November/
December, 1972), pp. 7?6-'7B4.
13. Hamilton, "Productivity," specifically commends the utilization of quantitative mea-
sures ". . . where output is very hard to measure . . . to improve the deployment of resources
so as to maximize the probability that our resources will be available at the time and place they
are needed most." (p. 787). This may be useful for fire protection where the presence of fire
fighters is the critical aspect of service provision. But it cannot be adequate for street-level bu-
reaucracies when resource availability may not be related to service quality.
14. Consider the following paragraph from a collection of essays on productivity. Mark
Holzer, ed., Productivity in Public Organizations (Port Washington, N.Y.: Kennikat Press,
1976). p. 19
Admittedly, there is an unevenness to productivity measurement. Some measures are relatively sophisti-
cated, othen crude. But in the common absence of any yardstick of productivity, even crude infOrmation is of
value. At least it is a means of introducing systematic quantitative analysis into the decision-making process.
Once that precedent is established, incremental refinements will undoubtedly lead to more sophisticated
measures. Quantifications should only be attempted, however, if the organization has the qualitative and
technical capacity to interpret and apply data meaningfully.
15. If pay increases for workers and the cost of city services depend on productivity, then
productivity measurement and assessment obviously become highly political. For example,
New York City workers have sought to measure the size of productivity savings in terms of the
net savings or additional income to the city from higher worker output. Fiscal managers, how-
ever, argue that productivity savings should be assessed only in terms of lower total wages
resulting from the need for a smaller work force to IH'COmplish the job. See the New York Times,
March 26, .1977.
16. For a discussion of these elementary aspects of productivity see Nancy S. Hayward,
"The Productivity Challenge," Public Administration Review (September/October, 1976), pp.
544-550
17. See New York Times, October 22, 1976, p. A26.
18. For a discussion of some of these service-rationing practices see Richard Weatherley
and Michael Lipsky, "Street-level Bureaucrats and Institutional Innovation: Implementing
Special Education Reform," Harvard Educational Review, vol. 47, no. 2 (May, 1977), pp.
171-197
19. Similar observations can be made for the apparent desirability of other modem man-
agement perspectives. Consider the following view of recent management control develop-
ments. "Critics of both sunset and zero-based methods have warned that these seemingly neu-
tral procedures are in fact skewed toward hardware and away from human services. Fighter
planes, miles of highway, or water projects are easily quantified; mental health, adequate nutri-
tion, or family well8re are not." Ross Milloy, "Is Carter Serious about Reorganizing the Govern-
ment? Should He Be?" Working Papers (January/February, 1978), p. 28.
20. On crisis see Murray Edelman, Political Language: Words that Succeed and Policies
that Fail (New York: Academic Press, 1977), chap. 3
21. I put "real" in quotes because when a saving is real and when it represents a reduction

259
Notes
in governmental effort is an empirical and normative question. Sometimes crisis can force man-
agement to attend to costs so that real savings are discovered, e.g., energy conservation by
reducing unnecessary wattage in bulbs. But at other times a change is simply justified by calling
it duplication or waste reduction although it may not be.
zz. In part, promotion and retention in street-level bureaucracies are not based on the
quality of service provision because service provision is so difficult to measure. Hence surro-
gates for effective service provision, such as tenure and advanced training, often bearing little
relationship to worker effectiveness, are used extensively to reward and promote workers.
These are not generally contradicted by more appropriate service delivery measures. On pro-
motion in street-level bureaucracies see John Van Maanen, "Working the Street: A Develop-
mental View of Police Behavior," in Herbert Jacob, ed., The Potential for Reform of the Crimi-
nal Justice System (Beverly Hills, Calif.: Sage, 1974); David Goodwin, Delivering Education
Services: Urban Schools and Schooling Policy (New York: Teachers College Press, 1977), pp.
66--67'
z3. The phrase is from Donald H. Sweet, Decruitment: A Guide for Managers (Reading,
Mass.: Addison Wesley, 1975).
z4. Letter from Hanna B. Leibowitz to New York Times, September z8, 1976, p. 38.
zs. "Frozen Means You Don't Move: The Impact of Budget Cuts on People in Mas-
sachusetts Institutions" (Massachusetts Advocacy Center, 1978), pp. 4&--47.
z6. Perhaps the most neglected aspect of the fiscal crisis is the extent to which the firing of
public employees represents a reduction in one of the critical functions of big city govern-
ments-the provision of relatively secure and decent jobs. After expressing great alarm for
many months over the fiscal crisis the New York Times eventually recognized this consideration.
"The trouble is, the bureaucracy also consists of people. Thus the fiscally sound demand for
greater economy and efficiency in the municipal health care bureaucracy could lead to the
discharge of thousands of hospital workers. In the absence of alternative job opportunities, the
result would'be suffering and despair in minority communities-and a sharp increase in welfare
rolls." November g, 1976, p. 36. Frances F. Piven has written presuasively on the redistrib-
utive aspects of urban fiscal liberalism and stringency. See an account of her views in the Boston
Globe, December g, 1976, p. 8.
z7. "Teaching is becoming an old people's profession." See ''Levittown Loses Its Younger
Teachers in Trims," New York Times, May 13, 1978, p. 27.

Chapter 12

1. V. 0. Key Jr., "Politics and Administration," in Leonard D. White, ed., The Future of
Government in the United States (Chicago: University of Chicago Press, 1942), p. 100. Cited in
Jesse McCorry, Marcus Foster and the Oakland Public Schools: Leadership in an Urban Bu-
reaucracy (Berkeley, Calif.: University of California Press, 1978), chap. 1.
z. The interplay between national culture and the organization of work is explored in
Michel Crozier, The Bureaucratic Phenomenon (Chicago: University of Chicago Press, 1g64), a
comparative study of French bureaucracy; Ronald Dore, British Factory~apanese Factory:
Origins of National Diversity in Industrial Relations (Berkeley, Calif.: University of California
Press, 1973). Arthur Stinchcombe explores the relationship of organizational structures to their
environmental origins in "Social Structure and Organization," in James March, ed., Handbook
of Organizations (Chicago: Rand McNally, 1g65), pp. 142-194 (esp. pp. 153-16g). I am in-
debted to Charles Sabel of the Massachusetts Institute of Technology fur his thoughtful com-
ments on this topic. See his unpublished essay, "Wor!<ers and World Views."
In this chapter I elaborate on the postulate that there is reciprocity between the larger soci-
ety and bureaucratic institutions. It follows that in different national (or even subnational) cul-
tural settings, there will be manifest differences in bureaucratic organization.
3 These findings are drawn from Gabriel Almond and Sidney Verba, The Civic Culture
 Notes
(Boston: Little, Brown, 1g65); and Michael Banton, The Policeman in the Community (London:
Tavistock, 1g64), cited in Elihu Katz and Brenda Danet, eds., Bureaucracy and the Public (New
York: Basic Books, 1973), p. 33 See their summary discussion of bureaucracy and culture, pp.
31-42
4 Katz and Danet, Bureaucracy and the Public, p. 34
5 The dynamics of the dialectic of expansion and contraction in public service benefits are
treated in Frances F. Piven and Richard Cloward, Regulating the Poor (New York: Pantheon,
1971). See also Michael Lipsky, Protest in City Politics (Chicago: Rand McNally, 1970), chap. 2;
Murray Edelman, Political Language (New York: Academic Press, 1977), chap. 3
6. To my knowledge few studies have inquired into public agency behavior under condi-
tions varying with agency need to attract clients. For one essay that addresses this consider-
ation, see Michael Lipsky and Morris I..ounds, "Citizen Participation and Health Care: Prob-
lems of Government Induced Participation," journal of Health Politics, Policy and Law, vol. 1,
no. 1 (Spring, 1976), pp. 85-111.
7 For illuminating discussions of the role of social welfare programs, broadly conceived,
in contemporary American society, see James O'Connor, The Fiscal Crisis of the State (New York:
St. Martin's, 1973); Piven and Cloward, Regulating the Poor; Ira Katznelson, "The Crisis of the
Capitalist City: Urban Politics and Social Control," in Willis Hawley and Michael Lipsky, eds.,
Theoretical Perspectives on Urban Politics (Englewood Cliffs, N.J.: Prentice-Hall, 1976), pp.
214-229.
8. See Jeffry Galper, The Politics of Social Services (Englewood Cliffs, N.J.: Prentice-
Hall, 1975).
9 Consider also Anthony Downs' discussion in "Why the Government Budget is Too Small
in a Democracy," World Politics, vol. 12, no. 4 (July, 1g6o), pp. 541-56:3.
10. On national variations in welfare benefit levels and administrative organization see
Harold Wilensky, The Welfare State and Equality: Structural and Ideological Roots of Public
Expenditures (Berkeley, Calif.: University of California Press, 1975).
11. On the symbolic significance of public policies see Edelman, Political Language.
12. I have stressed throughout the buffer role of street-level bureaucrats, especially in
chaps. 5, 6, and 9 For further discussion see Katznelson, "The Crisis of the Capitalist City," and
James D. Thompson, "Organizations and Output Transactions," in Katz and Dane!, eds.,
Bureaucracy and the Public, pp. 191-211.
13. Morris Janowitz directs attention to ways in which bureaucracies help shape the clients
with whom they later interact in Social Control and the Welfare State (New York: Elsevier,
1976).
14. Ibid., p. 105. Janowitz asserts that this is generally characteristic of service bureau-
cracies in welfare states: "Whether one is dealing with the format of public housing or with wel-
fare services associated with family assistance programs and community development, the
overall effect on the process of socialization is to separate and in fact isolate the clients from the
larger social structure and to seek to treat their needs in a very fragmented fashion. While these
programs have eliminated the stark misery of oppressive poverty and the fear of starvation, they
contain strong built-in limitations that thwart self-esteem and competence among recipients."
15. For approaches to politics that assume the ubiquity of conflict see Ralf Dahrendorf,
Class and Class Conflict in Industrial Society (Stanford, Calif.: Stanford University Press. 196g);
William Gamson, "Stable Unrepresentation in American Politics," American Behavioral Scien-
tist (November-Decem ber, 1g68), pp. 15-21. See also Lewis Coser, The Functions of Social
Conflict (New York: Free Press, 1g64).
16. Margaret Levi, "Poor People against the State," Review of Radical Political Economics,
vol. 6 (Spring, 1974), pp. 7&-79.
17. For one example of public employees seeking improved services fur citizens see the ef-
forts of the Service Employees International Union to obtain better patient care and treatment
facilities at Boston City Hospital. Boston Herald-American, May 25, 1978, p. 7
 Notes

Chapter 13

1. See Murray Edelman, Political Language: Words That Succeed and Policies thot Fail
(New York: Academic Press, 1977), Chap. 4
2. For a natural experiment in which, because of a strike in 1g67, a city discovered whether
it needed the services of a type of street-level bureaucrat, see Arnold Weber, "Paradise Lost:
Or Whatever Happened to the Chicago Social Workers?" in Joseph Loewenberg and Michael
Moskow, eds., Collective Bargaining in Government (Englewood Cliffs, N.J.: Prentice-Hall,
1973).
3 Fred Hechinger, "Smaller Classes Found to Produce Subtle Changes," New York Times,
April 10, 1979. p. Cs. The relationship of class size to achievement may depend on the subject
taught. A rece!lt South Carolina experiment found signifieantly higher reading scores in classes
averaging 19.9 students compared to classes averaging 26.7 students. But no appreciable dif-
ferences were discovered in math scores. See New York Times, June 22, 1977, p. 35
4 Boston Globe, September 11, 1977. p. 10.
5 For the argument that work structure is more important than training in detennining
physicians' attitudes and the character of their practice see Eliot Freidson, Profession of Medi-
cine (New York: Dodd, Mead, 1974), chap. 5
Richard Weatherley reminds me in a letter that for current workers, training often serves the
latent function of providing a break from the job and is a fringe benefit contributing to worker
morale, whatever the effect on practice.
6. See John H. McNamara, "Uncertainties in Police Work: The Relevance of Police Re-
cruits' Backgrounds and Training," in David Bordua, ed., The Police (New York: John Wiley,
1967). pp. 16:3-252.
7 See Harold Wilensky, "The Professionalization of Everyone," American journal of Soci-
ology, vol. 70, no. 2 (September, 1g64); Amitai Etzioni, The Semi-Professions and their Organi-
zation (New York: The Free Press, 1g6g).
8. Usually, increased professionalization is associated with deference of the society toward
occupations, discretionary judgment, citizen trust, and reciprocal altruism, as well as the char-
acteristics mentioned here. However, in some cases the tenn simply denotes an occupational
group's increased adherence to accepted occupational nonns. Thus increased professionalism
among police may mean less discretion and greater confonnity to legal standards and nonns of
police conduct.
9 Deboralt Stone, "Professionals and Social Services," (unpublished paper, March, 1976);
Gideon Sjoberg et al., "Bureaucracy and the Lower Class," in Sociology and Social Re-
search, vol. so (1g66), pp. 325-337
10. On peer review in medicine see Freidson, Profession of Medicine, pp. 13711'.
11. Ronald Gross and Paul Ostennan, ed., The New Professionals (New York: Simon and
Schuster, 1972).
12. For a discussion of increased client participation in service bureaucracies (one generally
consistent with the arguments of this chapter), see Alan Gartner and Frank Reissman, The Ser~
vice Society and the Consumer Vanguard (New York: Harper and Row, 1974).
13. On the limitations of group practice in medicine relative to dominant professional
nonns, see Eliot Freidson, Doctoring Together: A Study of Professional Social Control (New
York: Elsevier, 1975).
 Notes

Chapter 14

1. James W. Davis Jr. and Kenneth Dolbeare, Little Groups of Neighbors: The Selective
Service System (Chicago: Markham, 1968).
z. Martha Derthick, New Towns in-Town (Washington, D.C.: Urban Institute Press,
197Z).
3 Jeffrey Pressman and Aaron Wildavsky, Implementation (Berkeley: University of Cali-
fornia Press, 1973). For the first stirrings of the field of implementation studies, see Austin
Ranney, "The Study of Policy Content: A Framework for Choice," in Ranney, ed., Political
Science and Public Policy (Chicago: Markham, 1968), pp. 3-Zl.
4 C. Eugene Steurle, "Financing the American State at the Tum of the Century," in
W.E. Brownlee, ed., Funding the Modem American State, 1941-1gg6: The Rise and Fall of
the Era of Easy Finance (Cambridge: Cambridge University Press, 1996), pp. 409-444.
S Iris J. Lav, Elizabeth McNichol, and Robert Zahradnik, "Faulty Foundations: State
Structural Budget Problems and How to Fix Them" (Washington, D.C.: Center on Budget
and Policy Priorities, May 17, zoos). available at: http://www.cbpp.orglfiles/s-17-0Ssfp.pdf
(accessed December 10, zoog).
6. Contracting for services with for-profit agencies may be an entirely different matter.
The logic of profit-driven provision of services would predict that management give priority
to increasing caseloads and demanding that workers streamline and routinize interactions
with clients. See, for example, Janice Johnson Dias and Steven Maynard-Moody, "For Profit
Welfare: Contracts, Conflicts, and the Performance Paradox," Journal of Public Administra-
tion Research and Theory, vo!. 17, no. z (zoo7), pp. 18g--zu. Such pressures exist in the
public sector, to be sure, but perhaps not so overtly or legitimized by the rationale of the
market. The discussion of contracting for services draws on Steven Rathgeb Smith and Mi-
chael Lipsky, Nonprofits for Hire: The Welfare State in the Age of Contracting (Cambridge,
Mass.: Harvard University Press, 1993).
7 For a discussion on professionalism as an autonomous source of accountability in or-
ganizations, see Carolyn J. Hill and Laurence E. Lynn Jr., Public Management: A Three-Di-
mensional Approach (Washington, D.C.: CQ Press, zoog), pp. 3zz-3zs.
8. This section draws on Michael Lipsky, "Revenues and Access to Public Benefit," in
Jorrit de Jong and Cowher Rizvi, eds., The State of Access: Success and Failure of Democra-
cies to Create Equal Opportunities (Washington, D.C.: Brookings Institution Press, zooS),
PP 137-147.
g. Elizabeth McNichol and Nicholas Johnson, "Recession Continues to Batter State
Budgets; State Responses Could Slow Recovery" (Washington, D.C.: Center on Budget and
Policy Priorities, zoog), available at: http://www.cbpp.orglfiles/g-8-o8sfp.pdf (accessed De-
cember 10, zoog).
10. Jon Honeck, "The Governor's Announcement on Taxes" (Cleveland, Ohio: The Center
for Community Solutions, October 1, zoog), available at: http://www.communitysolutions
.corn/images/upload!resources/IncomeTaxStatementlooiOg.pdf (accessed December 19, zoog).
11. New York Times, October g, zoog, p. Azs.
1z. A staple of public opinion research is that people generally like what government
does, but tend to have a poor opinion of government as such. See, for example, Meg Bostrom,
"By, or for, the People? A Meta-analysis of Public Opinion of Government," (New York:
Demos, zoos). available at: http://www.demos.org!pubs/ByOrForthePeoplezooso4z6.pdf (ac-
cessed December 10, zoog).
13. Paul Pierson, "From Expansion to Austerity: The New Politics of Taxing and Spend-
ing," in Martin Levin et a!., eds., Seeking the Center: Politics and Policymaking at the New
Century (Washington, D.C.: Georgetown University Press, zoo1), p. 66.
14. The primary exception in the United States has been the transformation of the basic
welfare program for dependent children from an entitlement program to a time-limited as-
sistance program emphasizing work effort.
 Notes
15. "Kaiser/Harvard Consumer Protection in Managed Care Smvey, December, 1997,"
Caring, vol. 17, no. 3 (March, 1998), pp. 1z-zo, available at: http://www.ncbi.nlm.nih.gov/
pubmed!101790ZO (accessed December 10, zoog). Generally, support for government initia-
tives to solve social and economic problems has been on the rise since zoo1, as has skepti-
cism toward elected officials. For a review before developments associated with the election
of President Obama and the economic collapse, see Trends in Political Values and Core At-
titudes: 1987-Z007 (Washington, D.C.: Pew Research Center for the Public & the Press,
March zz, zoo7), available at: http://people-press.orwreportl31z/trends-in-political-values
-and-core-attitudes-1987-ZOO? (accessed November zz, zoog).
16. Among other works, see Murray Edelman, The Symbolic Uses of Politics (Urbana:
University of Illinois Press, 1964).
17. For example, one can test whether providing a low income family with a housing
voucher improves its members' life chances. See James Rosenbaum and Stephanie DeLuca,
"Is Housing Mobility the Key to Welfare Reform?" (Washington, D.C.: Center on Urban and
Metropolitan Policy, The Brookings Institution, September, zooo), available at: http://www
.brookings.edu/-/media!Files/rc/reports/zooo/ogmetropolitanpolicy_rosenbaum/rosenbaum.
pdf (accessed October z3, zoog).
18. See Lizbeth Schorr, "Charities' Work Demands Flexible Evaluation," Chronicle of
Philanthropy, August zo, zoog, pp. 33ff.; also Michael Edwards, Small Change: Why Busi-
ness Won't Save the World (San Francisco: Barrett-Koehler, zoog).
19. In his inaugural address, President Obama emphasized that the role of government
should be approached pragmatically, and that government accountability is of the highest
priority. "The question we ask today is not whether our government is too big or too small,
but whether it works-whether it helps families find jobs at a decent wage, care they can af-
ford, a retirement that is dignified .... [T]hose of us who manage the public's dollars will be
held to account-to spend wisely, reform bad habits, and do our business in the light of
day-because only then can we restore the vital trust between a people and their govern-
ment." Obama Inaugural Speeches, January zo, zoog, available at: http://obamaspeeches.com
(accessed December 10, zoog).
zo. For a discussion that begins to treat the complexity of generalizing about account-
ability over the range of occupational roles and work environments that are encompassed by
the term street-level bureaucracy, see Peter Hupe and Michael Hill, "Street-Level Bureau-
cracy and Public Accountability,'' Public Administration vol. 85, no. z (zoo7), pp. z7g--zgg.
Zl. See pp. 3. xii, this volume. Some have even suggested that the book may be read as
unequivocally endorsing the discretionary behavior street-level bureaucrats exhibit. See, for
example, Janet Coble Vinzant and Lane Crothers, Street-Level Leadership: Discretion and
Legitimacy in Front Line Public Service (Washington, D.C.: Georgetown University Press,
ZOO?). P g.
zz. Evelyn Brodkin, "Bureaucracy Redux: Management Reformism and the Welfare
State," Journal of Public Administration Research and Theory, vo!. 17, no. 1 (zoo7), pp. 1-17.
z3. New York Times, April zg, zoog, p. A1g.
z4. See pp. 196--1g8, this volume.
z5. But even clerks can exercise discretion. See Saul Weiner et a!., "Rationing Access to
Care to the Medically Uninsured: The Role of Bureaucratic Front-Line Discretion at Large
Healthcare Institutions," Medical Care, vol. 4z, no. 4 (April, zoo4), pp. 306--31z.
z6. See Ian Taylor, "Discretion and Control in Education: The Teacher as Street-level
Bureaucrat," Educational Management Administration & Leadership, vo!. 35, no. 4 (zoo7),
pp. 555-57Z; Tony Evans and John Harris, "Street-Level Bureaucracy, Social Work and the
(Exaggerated) Death of Discretion," British Journal of Social Work, vol. 34 (zoo4), pp. 871-
895.
z7. Virginia Department of Motor Vehicles, "Virginia Performs," available at: http://
vaperforms.virginia.gov/agencyleveVsrc/ViewAgency.cfm ?agencycode= 154 (accessed October
z4, zoog).
zB. Celia Hagert, "Updating and Outsourcing Enrollment in Public Benefits: The Texas
Experience" (Austin, Tex.: Center for Public Policy Priorities, November, zoo6), available at:
http://www.cppp.orwfiles/3fCPPP_PrivReport_ %z8FS%zg.pdf (accessed October 11, zoog).
 Notes
zg. On the hidden costs to citizens of efforts to save money through what are advertized
as essentially administrative reform, see Michael Lipsky, "Bureaucratic Disentitlement in So-
cial Welfare Programs," Social Service Review, vo!. 58 (1984), pp. 1-17.
30. Greg Marston, "Employment Services in an Age of E-Govemment," Information,
Communication & Society, vol. g, no. 1 (February zoo6), pp. 83-103.
31. David Landsbergen, "Screen Level Bureaucracy: Databases as Public Records,"
Government Information Quarterly, vo!. 21 (2004), p. 25.
32. In this section I draw on Michael Lipsky, "The Paradox of Managing Discretionary
Workers in Social Welfare Policy," in Michael Adler eta!., eds., The Sociology of Social Secu-
rity (Edinburgh: Edinburgh University Press, 1991), pp. 212-228.
33 Evelyn Brodkin, The False Promise of Administrative Reform (Philadelphia: Temple
University Press, 1986).
34 Jerry Mashaw, Bureaucratic Justice: Managing Social Security Disability Claims
(New Haven, Conn.: Yale University Press, 1983).
35 Robert Behn, Leadership Counts: Lessons for Public Managers from the Massachu-
setts Welfare, Training, and Employment Program (Cambridge, Mass.: Harvard University
Press, 1991). This account also draws on my experiences as a consultant with the Massachu-
setts Department of Public Welfare from 1983 to 1988.
The department leaders resisted evaluation through random assignment, despite consid-
erable pressure to do so, primarily for two reasons. First, they believed in placing people in
jobs, and thus considered it wrong to deny some citizens access to important services. Sec-
ond, they believed that it would not be possible to isolate a true control group, because so
many ET activities consisted of statewide publicity and collective efforts to assist welfare re-
cipients. See Behn's extensive discussion of ET's fit with the standards of random assignment
experimentation, chap. g.
36. Ibid., chap. 1.
37 For an account of a welfare agency that did not put comparable management tools
in place and failed in transforming workers' orientations, see Marcia K. Meyers et a!., "On
the Front Lines of Welfare Delivery: Are Workers Implementing Policy Reforms?" Journal
of Policy Analysis and Management, vo!. 17, no. 1 (1gg8), pp. 1-22. Soeren Winter and col-
leagues in many papers have explored on an empirical basis the contributions of manage-
ment initiatives, public policy dictates, and other interventions to bringing street-level bu-
reaucrats' performance in line with policy directives in Denmark See, for example, Peter J.
May and Soeren Winter, "Politicians, Managers, and Street-Level Bureaucrats: Influences on
Policy Implementation," Journal of Public Administration Research and Theory, vo!. 19, no.
3 (zoog), pp. 453-476.
38. For the origins of CompSTAT, see Jack Maple, The Crime Fighter: Putting the Bad
Guys Out of Business (New York: Random House, zooo).
39 In many articles Robert Behn seeks to identify the core elements of the CompSTAT
approach for a range of public management reforms that purport to build on the model. See,
e.g., "Designing PerformanceStat," Public Performance and Management Review, vol. 32, no.
z (December, zoo8), pp. 206-235
40. For another instance of shaping workers' behavior in the direction of generosity, see
Robert Garot, "Bias Forged through Suspicion: The Housing Gatekeeper Reconsidered," in
Stacy Bums, ed., Sociology of Crime, Law, and Deviance, vo!. 6 (Greenwich, Conn.: JAI
Press, zoos), pp. 77-104.
41. See p. 15, this volume.
42. Lisa Dodson, The Moral Underground: How Ordinary Americans Subvert the Un-
fair Economy (New York: New Press, zoog). This behavior has been described as "sabotage"
of agency policy in John Brehm and Scott Gates, Working, Shirking, and Sabotage: Bureau-
cratic Response to a Democratic Public (Ann Arbor: University of Michigan Press, 1997).
43 Steven Maynard-Moody and Michael Musheno, Cops, Teachers, Counselors: Stories
from the Front Lines of Public Service (Ann Arbor: University of Michigan Press, 2003), p.
24.
44 Michael Hill and Peter Hupe, Implementing Public Policy (London: Sage Publica-
tions, zooz), p. 27.
Notes
45 "The law's emphasis needs to shift from applying sanctions for failing to raise test
scores to holding states and localities accountable for making the systemic changes that im-
prove student achievement." "Joint Organizational Statement on No Child Left Behind
(NCLB) Act" (Boston, Mass.: FairTest, zoo4), available at: http://www.fairtest.or!joint%zo
statement%zocivil%zorights%zogrps%zo10-z1-04.html (accessed December 10, zoog).
46. To take an example with significant implications, in some schools serving Native
American students study units focusing on the history and culture of Native Americans-
subjects critical to helping native youth understand and appreciate their heritage-have been
dropped by schools giving priority to test results. Presentation by Hon. Ernie St. Germaine,
Honoring Nations Symposium, Cambridge, Mass., September 18, zoog.
47 Gerald Grant and Christine E. Murray, Teaching in America: The Slow Revolution
(Cambridge, Mass.: Harvard University Press, 1999), p. z15.
48. Dorothy Anagnostopoulos uses a street-level bureaucracy perspective to analyze
Chicago teachers' responses to a new accountability agenda in "The New Accountability, Stu-
dent Failure, and Teachers' Work in Urban High Schools," Educational Policy vol. 17, no. 3
(July zoo3), pp. zg1-316.
49 Linda Darling-Hammond and Milbrey Wallin McLaughlin, "Investing in Teaching
as a Learning Profession," in Darling-Hammond and Gary Sykes, eds., Teaching as the
Learning Profession (San Francisco: Jossey-Bass, 1999), pp. 376-411; see also Ron Haskins
and Susanna Loeb, "A Plan to Improve the Quality of Teaching in American Schools" (policy
brief, The Brookings Institution, Spring, zoo7), available at: http://www.brookings.edu/-/
media!Files/rc/papers/zoo7/spring_childrenfamilies_haskins/spring_childrenfamilies_
haskins.pdf (accessed November zg, zoog).
so. Mary E. Dilworth and Joseph A. Aguerrebere, "NCLB's Highly Qualified Teacher: A
Placeholder Definition," National Journal of Urban Education and Practice, vol. 1, no. z
(Fall, zoo7), pp. 111-135.
51. See Dilworth and Aguerrebere, p. 1Z7.
sz. Dilworth and Aguerrebere, p. 119.
53 See chap. 4
54 Letter to the legislature, Lewis H. Spence, Commissioner of the Department of So-
cial Services, April z4, zooz, p. 3
55 Lewis H. Spence, Letter to Fellow Employees, January z1, zoos, p. z.
56. Linda T. Kohn eta!., eds., To Err Is Human: Building a Safer Health System, Com-
mittee on Quality of Health Care in America (Washington, D.C.: National Academy Press,
zooo), available at: http://www.nap.edu/openbook.php?isbn=o3ogo68371 (accessed October
19, zoog). See pp. zog-z10, this volume.
57 On the relationship of citizen trust and experiences with street-level policy, see Bo
Rothstein, "The State and Social Capital: An Institutional Theory of Generalized Trust,"
Comparative Politics (July zooS), pp. 441-459.
58. See Margaret Levi, "A State of Trust," in Valerie Braithwaite and Levi, eds., Trust
and Governance (New York: Russell Sage Foundation, 1998), pp. 77-101.
59 Murray Edelman, Constructing the Political Spectacle (Chicago: University of Chi-
cago Press, 1988).
6o. See p. 5, this volume, and "U.S. Census of Government Employment by Geography
and by Government Function," March, zoo7.

266
 INDEX
Absenteeism, 17, 80, 143, 257n
Accountability, 159-165; and client
organization, 189; in education, 233;
increasing, 195; in medicine, 258n; and
performance measures, 165-169; and
policy making by street-level bureaucrats,
221-223;andproductivity, 170-172;and
professionalism, 203; and responsiveness,
230-231
Administrative law judges (ALJs), 226
Adoption agency workers, 130
Adult continuing education, 92
Advocacy, 72-75, 99; discouragement of, 185;
in the professions, 203
Affirmative action, 135-136
Aguerrebere, Joseph A., 232-233
Aid to Families with Dependent Children, 94
Alcoholism, 18, 20, 102, 112; among police,
32; and the ideology of benign
intervention, 119
Alex, Nicholas, 253n
Alford, Robert, 245n
Alienation, 17,75-80
Allen, Vernon,246n-247n
Almond, Gabriel, 260n
Alternatives to incarceration, 19-20, 43
Altruism, 71-73, 144, 185; and
professionalism, 201, 204
Ambulance services, 16, 136, 167
Anderson, James G., 247n
Antigovernment perspective, 217-221, 264
Apathy, 17
Appeals, 134-136
Appointments, 96, 98
Argyris, Chris, 241-242n, 250n, 257n
Aronson, Elliot, 246n
Arrest rates, 51-52, 114
Atkins, Charles, 226
Attrition, 174-175
Automated service systems, 224-225
Autonomy of street-level bureaucrats, 221-
225
Baker, Alan, 239n-240n
Banton, Michael, 261n
Barnard, Chester, 246n
Bar-Yosef, Rivka, 256n-257n
Becker, Howard, 155, 244n, 255n, 257n-258n
Behn, Robert, 227
Bell, Daniel, 240n
Bellow, Gary, 246n
Bendix, Reinhard, 251n
Bennett-Sandier, Georgette, 244n
Berger, Peter, 248n-249n, 251n
Bias, 141-142, 155-156; and bureaucratic
work settings, 85; and client
differentiation, 108-116; media emphasis
on government failures over successes,
218; racial, ll1, ll5, 181-182
Bittner, Egon, 256n
Blau, Peter, 51, 166, 240n, 247n, 250n, 258n
Blind, the, 109
Bloch, Alfred M., 32, 244n
Blum,Richard,244n
Bordua, David, 240n, 244n-245n, 247n,
249n,254n,256n
Boston, 10, 127-128, 135, 261n; Little City
Halls program in, 35, 92; police in, 246n;
pretrial diversion programs in, 20; public
housing in, 21-22, 65, 96, 101, 108
Boston City Hospital, 261n
Boston Housing Authority (BHA), 21-22, 108
Bowles, Samuel, 240n
Bridges, Gary, 247n
Brim, Orville, 247n, Jr.
Brintnall, Michael, 252n
Brodkin, Evelyn, 222
Budgetary constraints, 214, 216-217
Building inspectors, 120, 204; see also
Housing inspectors
Bureaucracies: reform of, 192-2ll; and
societal values, 180-188; see also
Organizations; Public service workers;
Street-level bureaucrats
Bush, George W, 231
Carey, Hugh, 240n
Carlin, Jerome, 130, 255n, 257n
Carr, Edwin, 244n, 255n, 257n
Cheating, 17
Chicago, 5, 99
Chickering, A. Lawrence, 240n
Child abuse and neglect, 155
Child Protective Services (CPS), 74-75, 233-
236
Cicourel, Aaron V., 247n, 250n, 252n
Citizen patrols, 194
Citizen review boards, 124
Index
Civil-rights compliance officers, 14
Civil rights movement, 212-213
Civil service system, 17, 23-24, 79, 82, 143,
206
Clark, Kenneth, 249n, 258n
Clerks, 14, 128-130, 239n
Clients, 83, 243n; championed by organized
public employees, 190-191; conceptions
of, held by public service workers, 141-
142, 151-156; control of, with procedural
routines, 117-128, 185; differentiation of,
105-116; and emergency procedures, 136-
139; impact of contact with, 190;
increasing autonomy of, 193-196;
increasing participation of, 208; informal
selection of, 102-103; nonvoluntary, 54-
59; organization by, 119, 189; psychological
burdens home by, 93-94, 104; referrals of,
132; screening of, 128-131; social
construction of, 59-70
Cloward, Richard, 10, 240n, 252n, 256n-257n
Cohen, David K., 248n
Collective bargaining, 80, 206; for client
needs, 190-191; see also Unions
Community action agencies, 10-11, 92
Community dispute resolution, 194
Community relations officers, 135-136, 146
Comprehensive Employment and Training
Act (CETA), 178
CompSTAT initiative, 228-229
Computers, 198
Conservative movement, 215, 217-221
Contracting for public services, 215-216, 224,
263n
Coping behavior: client differentiation as,
151-156; and client-processing mentality,
140-156; necessity of, 187; procedural,
121-125; routines as, 85; see also Routines
Correctional facilities, 12-13, 15, 147;
contradictory goals of, 41-42, 165; the
medical model in, 148; work conditions in,
31-32; see also Prisons
Corruption, 63
Coser, Lewis, 261n
Court clerks, 90, 94, 252n
Courts, 3, 5, 203; access to, 102; as agents of
social control, 11; bias in, 113-114; case
priorities in, 135; function of setting in,
117-118; goal conflicts in, 42;
rubberstamping of recommendations to,
129-131; service rationing by, 90; waiting
imposed by, 97-98; see also Judges
Couzens, Michael, 247n, 259n
CPS (Child Protective Services), 74-75, 233-
236
Creaming, 49, 107-108, 166
Creativity, 250n
268
Cressey, Donald, 257n
Criminal justice system, 15; see also
Correctional facilities; Courts; Prisons
Criticism of public services, 217-221
Crozier, Michel, 257n, 260n
Cumming, Eleanor, 248n, 250n
Cynicism, 78
Dahl, Robert, 252n, 254n
Dahrendorf, Ralf, 241n-2242n, 261n
Danet, Brenda, 248n-249n, 251n, 253n,
256n-258n, 261n
Davis, James W., 214, 256n
Davis, Karen, 251n
Decentralization, 116, 207-208
Decruitment practices, 176-177
Demonstration projects, 187
Denver Plan, 200
Department of Health, Education, and
Welfare, 164, 241n
Department of Housing and Urban
Development, 21
Derelicts, 18
Deutscher, Irwin, 254n-255n, 258n
Diagnostic related groups (DRGs), 225
Dilworth, Mary E., 232-233
Direct service provision, abandonment of,
215-216
Disability insurance determinations, 226
Discretion in policy implementation, 13-25,
84-85,149-150,221-229,233-236
District attorney, 22-23
Doctors, 73-74, 110, 254n; appointments
with, 98; emergency room, 130; and home
visits, 120; and the medicalization of social
problems, 148; Medicare claims by, 45; as
policy makers, 19-21,25;practice
standards of, 202-203; role conflicts of,
162; use of emergency procedures by, 138;
in the VA hospital system, 20-21, 23; see
also Medical practice
Dodson, Lisa, 230
Dolbeare, Kenneth, 214, 251n, 256n
Domestic relations courts, 130
Downs, Anthony, 243n, 245n, 251n, 257n,
261n
Draft boards, 84, 256n
Drug laws, 22
Drug treatment centers, 43, 93, 152
Easton, David, 251n-252n
Economic Opportunity Act of 1964, 41
Edelman, Murray, 219, 236, 249n-250n,
252n,258n-259n,261n-262n
Education, 5, 77, 206, 231-233; see also
Schools; Special education; Teachers
Educational vouchers, 93, 104
 Index
Eisenstadt, S. N., 248n-250n Goodrich, C. H., 245n
Eligibility, 60--61 Goodwin, David, 260n
Ellis Island, 62 Government's role in society, increased
Emergencies, 136--39 criticism of, 214-215, 217-221
Emergency rooms, 33, 37, 43, 66; bias in, Great Britain, 180
109; and client education, 139; difficulties Great Recession (2008), 217
of performance measurement in, 168; Greenberg, Stanley, 250n
resource husbanding in, 125-126; role of Greenlick,~erwyn,245n
clerks in, 129--130; treatment of clients in, Gross, Ronald, 262n
85; triaging in, 106; waiting in, 98--99; see Groub~Barbara,239n-240n
also Hospitals Gypsies, 124-125
Emerson, Robert, 248n-249n, 253n
Employment agency, 51 Hahn, Harlan, 240n, 249n
Employment and Training Choices (ET), Halfway houses, 67
226--228,265n Hamilton, Edward K., 259n
Employment counselors, 74, 107--8, 166 Handler, Joel, 249n-250n, 255n
Employment programs, 11, 226--228 Hansen, Donald, 243n
English, Richard, 246n, 248n, 25ln, Hasenfeld, Yeheskel, 246n, 248n, 25ln, 253--
253n-255n 255n
Epstein, Irwin, 256n Hatry, Harry, 259n
Equal opportunity officers, 136 Hawley, Willis, 40, 239n, 246n-248n,
Etzioni, Amitai, 76, 243n, 246n, 248n, 250n, 250n-25ln,255n,26ln
257n, 262n Hayward, Nancy S., 259n
Exceptional children, 70 Head Start, 7
Health care, 32, 72, 207; home, 194-195
Farrar, Eleanor, 248n Health centers, 10; neighborhood control of,
Favoritism, 151-152 196; queue systems in, 96
Fire departments, 136 Health service careers, 3, 5, 206
Fiscal crisis, 172--179, 186, 240n; and Hechinger, Fred, 24ln, 262n
bureaucratic accountability, 159; and the Henry, Jules, 254n
firing of public employees, 7--8, 260n Heumann, ~ilton, 245n
Fiscal stringency, 214, 216--217 Hill, ~ichael, 231
Fogelson, Robert, 245n Hirschman, Albert 0., 52, 247n
Food stampJ.rogram, 45, 89, 246n Hollingsworth, ~ary Jane, 249n-250n, 255n
Ford, Cera! , 246n Holzer, ~ark, 259n
Forer, Lois, 239n, 254n Home health care, 194-195
Forms, 120-21, 150 Home visits, 120, 150
For-profit agencies, public service contracts Hospitals, 114, 203; emergency procedures
with, 263n in, 136; and the ideology of benign
Freidson, Eliot, 112, 141, 248n-249n, 254n, intervention, 119; inquiry policies in, 209--
256n-258n, 262n 210; protocols for medical priorities in,
Fuchs, Victor, 240n l98;VA,20-21,23, l29,24ln;voluntary,
Funding, 92--93 109; see also Emergency rooms
Hosticka, Carl, 120, 240n, 243n, 245n,
Galper, Jeffry, 132, 239n-240n, 253n, 256n, 253n-255n, 257n
26ln Housing, 14,21-22,88, 135, 254n;
Gamson, William, 26ln discriminatory decision making in public,
Gartner, Alan, 240n, 262n 108; and emergency status, 137; inspection
Gaylin, Willard, 254n of, 52 (see also Housing inspectors);
Ginsburg, Sigmund, 242n integration of public, 46; and rent strikes,
Gintis, Herbert, 240n 63--64; role of applications clerk in public,
Glaser, Barney, 253n 129; selective aid to applicants for public,
Goal clarification, 164 65,97,137, l50-l5l;tenantmanagement
Coffman, Erving, 66, 154, 248n-249n, 254n, of public, 193
258n Housing inspectors, 52, 101, 127-128; see
Goldbricking, 143 also Building inspectors; Housing
Goldstein, Herman, 247n Human relations staffs, 135--136
Index
Human services, 27, 192-211; see also Public
service workers
Hupe, Peter, 231
Idealism; see Altruism
Ideology, 147-148
Immigration regulations, 113
Implementation studies, 213--214
Income inequalities, 7,54-55
Inflation,39, 170,176
Information, 90--94
Information technology, 224-225
Integration,21-22,46,246n
Isolation, work in, 203, 206, 208
Israel, 62, 108
Jackson, Kenneth T., 245n
Jacob, Herbert, 244n, 246n, 260n
Jacobs, Paul, 243n
Jacobson,Lenore,240n,249n
Jails, 5; see also Correctional facilities; Prisons
Janowitz, Morris, 261n
Japha, Anthony, 257n
Jargon, 90
Job satisfaction, 140
Job stresses, 141-144; and defenses against
discretion, 149-150; and specialization,
146; see also Role conflicts
Job training counselors, 153
Job training programs, 49, 227-228, 265n
Judges, 3, 7, 97, 145; and advocacy, 74; basis
of sentencing by, 109, 111-112, 124-125;
controlling administrative law judges, 226;
and the courtroom setting, 118; decision
making by, 128; insulation of, from client
demands, 100; job reconceptualizations by,
150--151; as policy makers, 13, 19-20, 25,
85; professional identity of, 190;
psychological sanctions from, 94; rubber
stamping by, 129-131; and their
nonvoluntary clients, 58, 62, 64, 68; work
conditions of, 30, 34, 44; see also Courts
Juryimpaneling,98
Kah, Marianne Stein, 239n
Katz,Eiihu,248n-2150n,256n-258n, 261n
Katznelson, Ira, 240n, 248n, 261n
Keith-Lucas, Alan, 249n-250n, 252n
Kelner, Bernard G., 244n, 255n
Kettleson, Jeanne, 246n
Key, V 0., Jr., 180, 260n
Kirkham, George, 32, 244n
Kirp, David, 245n, 250n, 253n, 255n
Kitsuse, John 1., 247n, 250n, 252n
Kozol, Jonathan, 258n
Krugman, Paul, 217
Kuh, Richard, 22-23
Landau, Martin, 40, 246n
Landlords, 135
Lasswell, Harold, 251n
Law enforcement personnel; see Police
officers
Law practice, 202, 206; see also Lawyers
Lawyers, 19, 65, 98, 206; accountability of,
163; client control by, 120--122, 124;
insulation of, from client demands, 100;
perceived need for, 91; practice standards
of, 202-203; psychological sanctions
imposed by, 94; role conflicts of, 42, 162;
and specialization, 146; training of, 71;
work conditions of, 30--31, 34, 36; see also
Legal services lawyers; Legal services
programs; Public service lawyers
Lazarus,Richard,251n
Lebeaux, Charles, 257n
Legal services lawyers: client control by, 120--
122, 124; routines of, 150; support for, 205;
see also Lawyers; Legal services programs;
Public service lawyers
Legal services programs, 96, 99;
neighborhood control of, 196; small group
decision making for, 207; see also Lawyers;
Legal services lawyers; Public service
lawyers
Leibowitz, Hanna B., 260n
Levi, Margaret, 240n, 249n, 261n
Levin, Martin, 44, 243n-244n, 246n-247n
Lewin, David, 239n
Libraries,5,58, 102
Liebow, Elliot, 252n
Life insurance industry, 84
Lindblom, Charles, 251n
Lindzey, Gardner, 246n
Lipsky, Michael, 239n-240n, 245n-246n,
249n-250n,252n,256n-257n,259n,2 61n
Little City Halls programs, 35, 92
Loewenberg,J.Joseph,240n,262n
Long Island Expressway, 33, 38, 199
Los Angeles, 32
Lounds, Morris,246n,252n,261n
Luckman, Thomas, 248n-249n, 251n
Lyford, Joseph, 243n
Mainstreaming, 77
Managers, 18--25, 242n
Mandatory sentencing legislation, 151
Manuals, 162
March, James, 242n, 251n, 257n, 260n
Margolis, B. L., 244n, 247n
Marriage counselors, 130
Marston, Greg, 224
Massachusetts, 177; alternatives to
incarceration in, 19-20; special education
issues in, 74, 127

Mayhew, Leon, 252n
Maynard-Moody, Steven, 230
McCorry, Jesse, 260n
McNamara, John H., 244n, 256n, 262n
Mechanic, David, 24, 242n, 249n, 254n
Media, emphasis on government failures over
successes, 218
Medicaid, 88
Medical examiners, 130
Medicalization of social problems, 12, 148
Medical practice, 120, 160, 202, 258n; neglect
and abuse in, 56; peer review in, 262n; see
also Doctors; Hospitals; Nurses
Medical review boards, 160
Medicare, 45, 69, 89, 225
Mediocrity, 144, 169
Mental health centers, 33-34, 92-93, 96;
client differentiation in, 112; client-staff
interaction in, 94
Mental health workers, 85
Mental hospitals, 66-67, 78, 85, 154; under
budget freeze, 177; commitment to, 130
Mental illness, 69
Mental retardation, 77
Mercer, Blaine, 244n, 255n
Merit increases, 206
Merton, Robert, 249n
Messinger, Sheldon, 243n
Methadone, 22
Mileski, Maureen, 112, 244n, 254n-255n
Milloy, Ross, 259n
Minorities, 10, 135, 204
Morris, FrankL., Sr., 24ln
Moskow, Michael H., 240n, 262n
Moynihan, Daniel P., 246n
Museums, 102
Musheno, Michael, 230
National Association for the Advancement of
Colored People (NAACP), Ill
National Board for Professional Teaching
Standards (NBPTS), 232
New York City, 7, 134-135, 137, 167; Board
of Education, Ill; 911 system in, 35;
police in, 32, 240n; rent-strike movement
in, 63--64; and the Stavisky bill, 240n;
teacher staff reductions in, 176-177;
Transit Authority in, 171; welfare rolls in,
22,91,103-104
Niederhoffer, Arthur, 247n, 255n
911 system, 35, 92, 129
Nivola, Pietro, 242n, 247n, 249n, 253n, 255n
Nixon Administration, 164
No Child Left Behind (NCLB), 231-233,
266n
No-fault case review for CPS, 235
Noncooperation, 17
Index
Nonprofit organizations, public seiVices from,
215-216
Nordlinger, Eric, 242n, 245n, 252n,
257n-258n
Nurses, 19, 110, 201-202; see also Medical
practice
Nutt, Thomas E., 255n
Obama, Barack, 220, 264n
O'Connor, James, 240n, 26ln
Olson, David J., 245n
Organizations: accountability in, 159--164;
goal clarification in, 164-165; performance
measures in, 165-169; and policy making,
82-86; see also Bureaucracies
Orren, Karen, 25ln
Osterman, Paul, 262n
Outsourcing of public seiVices, 215-216, 224
PapeiWork, 163, 244n, 258n; and use of
forms, 120-121, 150
Paraprofessionals, 116, 171, 194, 245n; legal,
198
Parole personnel, 12, 110
Paternalism, 79, 197
Peer review, 203, 262n; and peer support,
206-207,209
Pension rights, 143
Performance measures, 48-53; and
accountability, 162-163, 165-169, 199; and
merit increases, 206
Perlman, Robert, 100, 244n, 253n,
256n-257n
Perry, David, 24ln, 256n
Personality development, 66-70
Peter Principle, 126
Physicians; see Doctors
Pierson, Paul, 217
Piliavin,IIVing,244n,247n,249n,254n
Pilot programs, 187
Piven, Frances Fox, 10, 240n, 252n, 256n,
260n-26ln
Police officers, 239n-240n, 257n;
accountability of, 159, 163-169; advocacy
by, 74; as agents of social control, ll-12;
alienation of, 78-79; allegations of
brutality by, 135; altruism of, 185;
American, compared to British, 180; and
bias, 108-111, 114; and citizen patrols,
194; and citizen renew boards, 124; client
control by, 122-125, 136; CompSTAT
initiative, 228-229; as deliverers of
government policy, 3-6; and department
reform, 228-229, 247n; emergency
procedures of, 136-137; and the fiscal
crisis, 175; incompetent, 187; insulation of,
from client demands, 100; manipulation of
Index
Police officers (cont.)
statistics by, 259n; performance measures
for, 50-52; as policy makers, 13-15, 18--19,
25, 84-86; private goal definition by, 145-
146; professional norms of, 190; reality-
based training for, 209; resource
husbanding by, 125-126; role conflicts of,
40-48,73-75,105, 115,246n-247n;
rubber-stamping of recommendations by,
129-131; self-image of, 82, 114; service
rationing by, 87-89, 92, 95, 102-103, 106;
small group decision-making by, 207;
support for, 205-206, 210; and their
nonvoluntary clients, 54, 62-63. 66; and
their partners, 203-204; work conditions
of, 30--32, 34, 39; work rewards of, 75
Police review boards, 124, 135, 160
Policy making, by street-level bureaucrats,
13-25,84-85,207,221-2 23
Post office, 88
Preschool programs, 7
Pressure specialists, 133-135
Pretrial diversion programs, 20
Principals, 242n
Prisons, 5-6,57, 114, 154; alternatives to,
19-20; client control in, 117, 119, 256n;
and corrections policy, 12; policy making
in, 13-14; and stigma, 69; work conditions
in, 78; see also Correctional facilities
Privacy, 64, 118
Private sector, provision of public services by,
215-216,224
Probation officers, 20, 125; weight of
recommendations by, 25, 130-131
Productivity, 170-172
Professionalism, 189-190, 262n; ideology of,
147-148; and the myth of service altruism,
71-80; prospects and problems of, 201-211
Prosecutors, 150-151
Prottas, Jeffrey, 242n-243n, 248n,
252n-253n,255n
Psychiatrists, 20, 31, 44; client differentiation
by, 111
Psychological sanctions, 93-94
Psychologists, 5, 77, 111, 113
Public defenders, 74, 101, 112, 145;
rubberstamping by, 130; working
conditions of, 29, 34
Public housing; see Housing
Public service lawyers, 3-7; altruism of, 185;
bias among, 113; as teachers of the client
role, 60-62, 69; who quit, 143; working
conditions of, 29-30, 34, 36; see also
Lawyers; Legal services lawyers; Legal
services programs
Public service workers: accountability of,
159-165, 170-172; advocacy by, 72-75;
272
alienation of, 75-80; and bureaucratic
reform, 192-211; client control by, 128--
139; client differentiation by, 105-116; and
the client-processing mentality, 140-156;
collective bargaining by, for client needs,
190-191; deliverers of government policy,
3-4, 8--12; evolving role of, 212-214,236-
237; and fiscal crisis, 172-179; husbanding
of resources by, 125-128; and the myth of
service altruism, 71-80; noncooperation
of, 17, 23-25; performance measures for,
165-169; as policy makers, 13-25, 144;
reductions in during recession, 217; role
conflicts of, 188--191; routines and
simplifications of, 83-86; service rationing
by, 87-104; and societal values, 180-188;
supporting idealism in, 204-211; work
conditions of, 27--39, 81-83; see also
Street-level bureaucrats
Pynoos,Jon,242n, 249n
Quality control in welfare administration,
225-226
Queuing, 95-99
Quinney, Richard, 240n, 254n
Quitting, 17
Racism, 111, 115, 181-182,213
Raffel, Jeffrey, 246n
Rainwater, Lee, 251n
Rape, 110-111
Rationingservices,87-10 4, 173;through
client differentiation, 105-116
Receptionists, 128--129, 198, 254n
Record keeping, 163, 258n; see also
Paperwork
Red tape, 63
Reference groups, 45-48, 57, 81; promoting
client impact on, 208
Referrals, 132-133
Reich, Charles, 240n
Rein, Martin, 240n, 246n
Reiss, Albert, Jr., 136, 241n, 247n, 252n, 256n
Reissman, Catherine Kohler, 96, 244n-245n,
253n
Reissman. Frank, 240n, 262n
Rent-strike movement, 63-64
Research standards for public services
evaluation,219-220
Resources, deployment of: and emergency
procedures, 138; husbanding of, 125-128;
by public-service workers, 167; results of
added, 199-200
Responsiveness of public services, 229
Retirement, 143
Review boards, 124, 135, 160
Richardson, James, 245n

Riley, Judy, 250n, 255n, 258n
Rist, Ray C., 67-68, 249n, 255n
Rockefeller, Nelson, 22
Rogers, David, 251n
Role conflicts, 40-48, 105, ll5; and advocacy,
73-75;forpollce,246n-247n
Rosenthal, Robert, 240n, 249n
Rosett, Arthur, 257n
Rossi, Peter, 254n, 258n
Roth,Jullus,85,248n, 251n,253n,255n
Routines, 83-86, 251n; and bureaucratic
reform, 199; and client control, 121-125;
consequences of, 133-136, 139; pollee,
120, 122-125;andservicerationing,99-
104; worker attachment to, 149
Roxbury Multi-Service Center, 100, 132
Rubber-stamping, 128-131
Rubinstein, Jonathan, 242n, 247n, 252n, 255n
Ryan, William, 257n
Sabel, Charles, 260n
Salaries, 5-7
Sanitation services, 167, 173
Sarbin, Theodore, 246n-247n
Savas, E. S., 242n, 247n
Scheff, Thomas, 243n
Schelling, Thomas, 248n
Schild, E. 0., 256n-257n
School counselors, 73
School psychologists, 62
Schools, 98, 146-147; accountability in, 159,
165; bias in, 108, 1ll-ll4; client control
in, ll9, 123-124; decentralization of, 160;
Denver Plan for, 200; expulsion from, 88;
function of settings in, ll8; goal conflicts
in, 40, 42-44, 46, 165; the medical model
in, 148; neighborhood control of, 196;
nonvoluntary clients of, 54-56; pressure
specialists in, 133; small group decision
making in, 207; socialization function of,
la--ll; specialization in, 77, 178; tracking
in, 63, 67-70, 106, ll2, 130; use of
community resources by, 194; see also
Education; Special Education; Teachers
Schultz, Stanley, 245n
Schwartz, Barry, 244n, 252n-253n
Schwartz, Charlotte Greene, 256n
Schwartz, Morris, 256n
Schwartz, Richard, 249n-250n
Scott, Robert, 254n
Screening, 128-129
Sears, David, 256n
Secretaries, 128-129
Seidman, David, 247n, 259n
Selective Service System, 84, 214, 256n
Self-actualization, 250n
Self-images, 66-67, 82, ll4, 183
Index
Semi-professionals, 194; see also
Paraprofessionals
Service Employee International Union, 261n
Service vouchers, 193-194
Settings, ll7-120
Sharkansky, Ira, 251n
Silver, Allan, 240n, 245n
Silver, Carol Ruth, 245n
Simon, Herbert, 242n, 251n, 257n
Simplifications, 83-86, 251n; and
bureaucratic reform, 199
Sjoberg, Gideon, 262n
Skolnick, Jerome, 249n-250n
Slowdowns, 143, 175
Snyder, John A., 255n
Social security benefits, 91-92
Social Security Disability Insurance (SSDI)
program, 226
Social workers, 3-7, 77, 105, 150; altruism of,
185; client conceptions of, 155;
discretionary roles of, 197,221-229, 233-
236; incompetent, 187; insulation of, from
client demands, 100; paternalism of, 79-
80; performance measures for, 50; as
policy makers, 19-20; professionalization
of, 190, 201-202; reality-based training for,
209-210; reporting required of, 182; role
conflicts of, 162, 183; rubber-stamping of
recommendations by, 13<f--131; site visits
by, 120; and specialization, 146; and their
involuntary clients, 57, 63, 65; work
conditions of, 3a--31, 36
Sornoff, Paula, 241n, 256n
Special education, 74, 77-78, 109, 127;
function of, 135; and issues of
specialization, 147
Specialization, 77-78; agency function of,
146-147; and fiscal crisis, 175; and
inequitable distribution of practitioners,
202
Spence, Harry, 233-234
Status, 202, 206
Stavisky bill, 240n
Stealing, 17
Steiner, Gilbert, 240n, 246n, 252n
Stereotypes, ll5, 140, 142, 155
Stevenson, Keith, 241n
Stigma,68-70,91-92, 182,184
Stinchcombe,Artbu~260n
Stone, Clarence, 258n
Stone,Deborah,245n,250n, 262n
Strauss, Anselm, 253n
Street-level bureaucrats: accountability to
policy, 221-229; as agents of social control,
8-12; conclusion, 236-237; contradictory
tendencies of, 188-189; defined, 3; as
deliverers of government policy, 3-4;
273
Index
Street-level bureaucrats (cont.)
evolving policy environment, 212--221;
goal conflicts of, 40-48; investing in, 229-
236; nonprofit organizations, 216; and
nonvoluntary clients, 54--59; performance
measures for, 48-53; policy making by,
13-25, 84--85,207, 221-223; scope of
services provided by, 4-8; shaping
performance of, 221-229; and the social
construction of clients, 59-70; work
conditions of, 27-39; see also Public
service workers; individual occupations
Sudnow, David, 112, 253n-255n
Supervisors, 242n
Support services, 31
Sweet, Donald H., 260n
Sykes, Gresham, 24ln, 244n, 248n
Teachers,93, 105,107-108, 198,247n;
accountability of, 163, 166-169; as agents
of social control, 11; alienation of, 78-80;
altruism of, 185; client conceptions of,
155, 254n; client control by, 120, 122,
124; client differentiation by, 110--111,
113, 152--153; collective bargaining for,
for better classroom conditions, 190--191;
as deliverers of government policy, 3-5,
7-8; under the Denver Plan, 200;
discretionary roles of, 197; and the fiscal
crisis, 175--177; incompetent, 187;
paternalism of, 79-80; performance
measures for, 50-53, 232--233; as policy
makers, 13, 15--16, 18-19, 85; private goal
definition by, 145--146; professionalization
of, 190, 201-202; reporting required of,
182; role conflicts of, 162, 183, 186-187;
rubber-stamping by, 130; self-image of,
82, 114; service rationing by, 89-90, 94;
socializing role of, 61--63, 67; and
specialization, 77; support for, 205--206,
210; and their nonvoluntary clients, 57-
58, 61--63; training of, 71; who quit, 143;
work conditions of, 30, 32, 36-37, 39; see
also Education; Schools
Teaching machines, 198
Team approach to CPS case work, 235
Tenure system, 93, 143
Thayer, Frederick, 250n
Thompson, Elizabeth J., 254n, 258n
Thompson, James D., 166-167, 248n, 255n,
258n,26ln
Thompson, Victor, 25ln, 257n
Time, 89-90
Tracking, 106, 112; effects of, 63, 67-70; and
rubber-stamping, 130
Trauma-team personnel, 109-110
274
Trees, Dennis, 255n
Triage, 106-107, 118, 138
Turnover, 143, 174
Ubell, Earl, 244n
Uniforms, 118
Unions, 7, 17, 24; and bureaucratic reform,
206, 210; and educational voucher
systems, 93; efforts by, for clients' needs,
80, 190--191
Universities, 39, 92--93, 208
Upward Bound program, 49-50, 107
Urban renewal programs, 8
U.S. Census Bureau, 5
Van Hom, Carl, 24ln-242n
VanMaanen, John, 244n, 255n, 260n
Van Meter, Donald, 24ln-242n
Verba, Sidney, 260n
Veterans Administration benefits, 92
Veterans Administration hospitals, 20--21, 23,
129,235,24ln
Violence, 31-32, 110--111, 135; and police
public image, 123
Voucherproposals, 93, 160,193-194
VVagefreezes, 175--176
VVaiting, 89, 93; in queues, 95--99; and triage,
106
VVar on poverty, 11, 41
VVaskow, Arthur, 245n, 250n
VVaste, 17
VVeatherley, Richard, 244n, 249n-250n, 255n,
257n,259n,262n
VVeber, Arnold, 262n
VVeinberg, Martha VVagner, 242n-243n
VVeiss, James, 245n
VVelfare agencies, 10, 46, 88, 139;
accountability in, 159, 164--165; and client
organizations, 119; discretionary options
in, 149, 197,222-223;emer gency
procedures in, 136; in New York City, 22,
91, 103-104; nonvoluntary clients of, 54-
55, 64, 66, 69; policy making in, 14; and
program goal conflicts, 41-43;
psychological burdens borne by clients of,
93-95, 104; quality control in, 225--226;
rationing of services by, 90; resource
husbanding in, 126-127; specialization in,
146; statistics generated by, 51; see also
VVelfare workers
VVelfare rights movement, 63, 149
VVelfare state, 8, 11-12, 183; contradictions
in, 189
VVelfare workers, 4; case presentations by,
134; client control routines of, 124-125,
 Index
150; discretion controls for, 225--226, Wilson, James Q., 229, 241n, 244n-247n,
265n; failure of, to differentiate among 250n-251n,257n-258n
clients, 122; response to 1996 reforms, Withdrawal behaviors, 142-144, 257n
222-223; self-image of, 82, 114; Wolf, June Grant, 252n
transforming orientations of, 226-228; see Wolff, Robert P., 240n
also Welfare agencies Women's rights movement, 135
Werthman, Carl, 244n, 249n, 254n Wynne, Edward, 258n
Westley, William, 244n
Wheeler, Stanton, 247n, 250n, 254n Yates, Douglas, 243n
White, Leonard D., 260n
Whitney, Richard, 256n Zald, M., 256n
Wildavsky, Aaron, 257n Zimmerman, Don, 243n-244n
Wilensky, Harold, 257n, 261n-262n Zoos, 102
Willemain, Thomas, 241n, 244n Zubkoff, Michael, 256n

275
STREET-LEVEL BUREAUCRACY
AWARDS FOR THE PREVIOUS EDITION
Winner of the 1980 C. Wright Mills Award, Society for the Study of Social Problems
Winner of the 1981 Gladys M. Kammerer Award, American Political Science Association
Winner of the 1999 Aaron Wildavsky Enduring Contribution Award of the Policy Studies Organization

First published in 1980, Street-Level Bureaucracy received critical acclaim for its insightful study
of how public service workers, wielding considerable discretion in how to execute their jobs,
function as policy decisionmakers. Three decades later, public urgency to bolster the availability
and effectiveness of healthcare, social services, education, and law enforcement has intensified,
making Street-Level Bureaucracy more relevant now than ever. In this thirtieth anniversary
expanded edition, Michael Upsky revisits the territory he mapped out in the first edition to reflect
on significant policy developments and show that street-level bureaucracies can be improved and
work in public service fields can be rewarding.
Street-Level Bureaucracy examines how discretionary services, work conditions, and work
practices interact to influence client outcomes. Street-level bureaucrats-from teachers and police
officers to social workers and legal-aid lawyers-interact directly with the public and so represent
the frontlines of government policy. Upsky argues that these relatively low-level employees in
human service agencies labor under huge caseloads, ambiguous agency goals, and inadequate
resources. When combined with substantial discretionary authority and their ability to interpret
policy on a case-by-case basis, the difference between government policy in theory and policy in
practice can become vast.
This seminal study tells a cautionary tale of how decisions made by overburdened workers
in underfunded government agencies translate into ad-hoc policy changes impacting peoples'
lives and life opportunities. This expanded edition of Street-Level Bureaucracy underscores that,
despite its challenging nature, street-level work can be a conduit for, rather than a barrier to,
providing services to citizens.

PRAISE FOR STREET-LEVEL BUREAUCRACY

"A rich, mature piece of scholarship ....An excellent book. "
-DOUGLAS YATES, American Political Science Review
"Provocative, well written, and full of marvelous insights into the service patterns
and practices of human services organizations .... A major contribution. "
-YEHESKEL HASENFELD, Socia/ Science Review
"A seminal study .. .. By far the most acutely observed and analytically
interesting work on this general subject."
-AARON WILDAVSKY
"Highly illuminating .... Provides valuable information on the interface between
the street-level human service bureaucrats and their clients."
-FRANK RIESSMAN, editor, Social Policy
"One of the most important recent books on urban affairs and administration. "
-Choice

ISBN-13: 976-0-67154-544-2
ISBN-10 : 0-67154-544-b
RUSSELL SAGE FOUNDATION
New York, New York I www.russellsage.org

10871111 f